14

Processability theory

Manfred Pienemann and Jörg-U. Keβler

Historical discussion

The basic ideas underlying Processability Theory (PT) developed as a result of the shortcomings of older approaches which predate PT. The goal of PT as well as its predecessor approaches is to explain second language (L2) acquisition based on the following logic: at any stage of development the learner can process only those L2 linguistic forms which the current state of the language processor can handle.

The conceptual development from the Multidimensional Model (MM) (Meisel et al., 1981) to Processability Theory (Pienemann, 1998a, 2005) has a 20-year history. There is a substantial difference between PT and ideas that precede it. Pienemann (1998a) took stock of the ideas predating PT, including the MM (Meisel et al., 1981), Clahsen's (1984) Strategies Approach, the Teachability Hypothesis (Pienemann, 1984, 1989) and the Predictive Framework (Pienemann and Johnston, 1987). PT was designed to overcome the limitations of these earlier approaches. Table 14.1 presents information about PT as well as earlier versions that led up to PT.

As illustrated in Table 14.1, the MM focused on the issue of determining developmental sequences in second language acquistion (SLA) corpora. One has to bear in mind that this line of research addressed some of the key issues discussed in the late 1970s and early 1980s which focused on the relationship between data and their proper description. The issue was whether every change found in an interlanguage (IL) constitutes evidence of IL development and if the changes that can be found need to be measured in native language terms. As the name suggests, the MM assumes that IL development is not linear and that instead ILs contain at least two dimensions. This assumption has repercussions for determining stages of development, and IL differences can be understood to represent either development or inter- or intra-learner variation. The MM utilizes implicational scaling and probabilistic rules to operationalize the emergence criterion for acquisition. The MM has been criticized for a lack of falsifiability. Larsen-Freeman and Long (1991) note that there is no independent motivation for variational IL features and that as a result there is a problem in not being able to falsify a developmental sequence because any deviation from a predicted sequence can be considered a variational feature. This shortcoming was rectified in PT (Pienemann, 1998a) which formally constrains development as well as variation in a predictable and testable manner as further discussed below.

The Strategies Approach was developed by Clahsen (1984) to explain the development of German L2 word order. It is based on a set of processing constraints which are assumed to be shed as IL development progresses. In this way, the strategies approach complements the MM by providing an explanation of the developmental dimension. Nevertheless, the MM and the Strategies Approach constitute two separate conceptual entities. Meisel (1983) produced his own version of the Strategies Approach that is more closely related to Slobin's(1973) and Andersen's (1984) acquisition strategies. Clahsen's strategies approach was subject to strong criticism (e.g., White, 1991). While some of these criticisms were refuted, the main limitations remained, including the problem with the model being restricted to constraints on L2 German word order and its undefined relationship to grammatical representation. These limitations were overcome in PT by utilizing a psychologically and typologically plausible theory of grammar (Lexical-functional Grammar [LFG], cf., Bresnan, 2001) that has been applied to a large variety of languages.

Table 14.1 PT and its predecessors and relatives

Model	Key references	Key concepts	Scope
Multidimensional Model	Meisel et al., 1981; Pienemann et al., 1993,	implicational scaling, probabilistic rules, emergence criterion, two dimensions in L2 dynamics: (1) development (2) variation	descriptive framework for dynamic processes in L2 development
Strategies Approach	Clahsen, 1984	development of L2 German word order determined by shedding of processing constraints	explanation of German L2 word order development
Predictive framework	Pienemann and Johnston, 1987	development of ESL morpho-syntax determined by incremental development of processing resources	Explains developmental patterns in ESL and GSL morpho-syntax
PT	Pienemann, 1998a, b, 2003, 2005; Pienemann et al., 2005	processability hierarchy modeled in linguistic theory (LFG), hypothesis space constrains development and variation (testable), factorization and other refined descriptive methods, partial transfer hypothesis (DMT), task variation: steadiness hypothesis, stabilization, L1-L2 differences, basis for L2 profiling	explains L2 development and variation in syntax and morphology cross-linguistically

The Teachability Hypothesis (Pienemann, 1984, 1989) was based on the MM and the Strategies Approach. It predicts that stages of SLA cannot be skipped through teaching intervention because of the implicational nature of the processing strategies. It also predicts that variational features are not subject to the same constraints on teachability. In other words, the Teachability Hypothesis defines constraints on teachability; however, it does not predict sufficient conditions for teaching to be successful. The Teachability Hypothesis was later incorporated into PT (Pienemann, 1998a).

All of the critical points listed so far gave rise to the development of Processability Theory, especially with regard to the lack of falsifiability in the MM and the inability of the Strategies Approach to link up to grammatical knowledge and its lack of typological plausibility.

PT was designed to address the two key issues of a theory of language acquisition, (i) the developmental problem (“why do learners follow universal stages of acquisition?” cf., Clahsen, 1992) and (ii) the logical problem (i.e., “how do learners develop linguistic knowledge—given that it cannot be inferred from the input without innate knowledge?” cf., Wexler and Culicover, 1980; Wexler, 1982; cf., VanPatten, Chapter 16, this volume). The 1998 volume on PT focused on the first issue. The 2005 volume also addressed the logical problem. This modular approach to explaining SLA was possible because of the incorporation of a grammatical theory (i.e., LFG, cf., Bresnan, 2001). This approach also permitted a clear and falsifiable differentiation between the two dimensions of SLA postulated by the MM. Both dimensions are constrained by the hypothesis space defined by PT.

Core issues

The basic idea

The basic idea underlying PT (Pienemann, 1998a, b, 2005) is the following: at any stage of development the learner can process only those L2 linguistic forms which the current state of the language processor can handle. Therefore understanding the architecture of the language processor and the way in which it handles a second language enables one to predict the course of development of L2 linguistic forms.

The architecture of the language processor accounts for language processing in real time and within human psychological constraints. Incorporating the language processor in the study of SLA therefore brings to bear human psychological constraints that go beyond the logico-mathematical considerations of learnability theory.

The core of PT is formed by a universal processability hierarchy that is based on Levelt's (1989) approach to language production and that is modeled using LFG (Bresnan, 2001). This universal basis permits PT to predict developmental trajectories for any second language. The notion “developmental trajectory” implies a developmental dimension known as “staged development” as well as a variational dimension accounting for individual differences between developmental trajectories as illustrated in Figure 14.1.

Figure 14.1 shows two different developmental trajectories, T1 and T2, which are based on the same set of developmental stages (indicated by the dotted horizontal lines). The two developmental trajectories differ with respect to the IL varieties that are developed at each stage (indicated by vertical lines).

In this paradigm, each stage corresponds to a set of grammatical rules that share certain processing routines, and each IL variety represents a specific variant of the grammatical rules.

Figure 14.1 Different developmental trajectories

For instance, in English Second Language (ESL) question formation the following developmental sequence has been found (e.g., Pienemann, 1998a):

Learners attempting to produce “Aux-second” before level 5 (i.e., before they are ready for this structure) have been found to produce the following interlanguage variants:

1. A   What she seen?
2. B   What has seen?
3. C   What she has seen?
4. D   She seen this?

Variants A to D have in common that they get around placing the auxiliary before the subject. In other words, they constitute different solutions to the same learning problem. In Figure 14.1 each of the different solutions is represented by a vertical line. In the course of L2 development, learners accumulate grammatical rules and their variants, allowing them develop an individual developmental trajectory while adhering to the overall developmental schedule. In this way, PT defines a two-dimensional space (i.e. “Hypothesis Space”). Both dimensions of Hypothesis Space are constrained by the processability hierarchy.

Key claims and constructs

PT entails the following key claims:

(a)   L2 development follows universal stages that are constrained by the processability hierarchy.

(b)   IL variation is limited and regular, and this limitation and regularity is caused by the constraints inherent in the processability hierarchy.

(c)   L1 transfer is constrained by processability. This implies that L1 forms can be transferred to the L2 only when they can be processed in the developing L2 system. In other words, this entails an operationalized “partial transfer” position (cf., Pienemann et al., 2005).

(d)   Task variation is constrained by the processability hierarchy. This claim results in the “Steadiness Hypothesis” which predicts that a specific IL will be placed at one and the same stage of development in different tasks as long as they refer to the same skill type.

(e)   Both, L1 and L2 acquisition is constrained by the processability hierarchy. Nevertheless, both types of acquisition may be associated with fundamentally different developmental trajectories.

(f)   Bilingual language development can be compared across different languages on a universal scale using the processability hierarchy inherent in PT (Pienemann et al., 2011).

These claims are based on two key constructs: (a) The processability hierarchy and (b) Hypothesis Space. The processability hierarchy is based on a universal set of processing resources that is modeled using LFG (Bresnan, 2001). In Pienemann (1998a) the hierarchy was based solely on the notion of transfer of grammatical information within constituent structure. The second theory module of PT (Pienemann et al., 2005) the processability hierarchy covers the relationship between conceptual structure and surface grammatical structure which is also modeled using LFG.

Figure 14.2 A simplified account of the processability hierarchy

A simplified account of the processability hierarchy (with a focus on constituent structure) is given in Figure 14.2 where three example constituent-structures are listed in the left-hand column. The second column specifies the type of information exchange possible at each stage. English morphological structures are given in the next column to exemplify the types of structures possible at each stage, and the information transfer involved in the generation of these structures is illustrated in the column on the right-hand side. For instance, for the noun phrase “two kids” (e.g., in the sentence “he has two kids”) the information “plural” has to be exchanged between the determiner and the noun; this information does not need to be exchanged with any other constituent in the sentence.

The notion “Hypothesis Space” was mentioned above in the context of developmental trajectories (cf., Figure 14.1). The structures that are processable at any given stage are constrained by the available processing resources. As can be seen in Figure 14.2, at the stage “phrase” grammatical information can be exchanged only within phrases, not beyond the phrasal boundary. Subject–verb agreement cannot be processed at the stage “phrase” because this structure requires grammatical information to be exchanged beyond the phrasal boundary. In other words, processing resources define and constrain the range of possible production grammars for every level.

At the same time, these constraints leave sufficient leeway for learners to find different solutions to structural learning problems. We illustrated this above with the example of the position of auxiliaries in English WH-questions. This position requires processing procedures at the sentence level in the hierarchy. L2 learners can nevertheless produce WH-questions before these procedures are processable. When they attempt to do this, learners have four structural options (variants A–D above) that avoid the placement of the auxiliary before the subject (i.e., in second position). The options available are all processable using the resources available at the previous stage, and the number of options is limited because of the limited resources that are available. The fact that learners need to circumnavigate a structural problem (here Aux 2nd) is caused by the constraints inherent in the hierarchy. In this way, possible and predictable developmental trajectories are constrained by the processability hierarchy.

PT principles and hypotheses

PT is based on four psycholinguistic principles and hypotheses which will be summarized below: (i) transfer of grammatical information, (ii) a lexically driven grammar, (iii) lexical mapping, and (iv) the TOPIC Hypothesis.

As mentioned above, the 1998 PT module focused on constituent structure and the exchange of grammatical information within it, using feature unification. The modeling of feature unification, as envisaged in this approach, is illustrated in Figure 14.2. In the sentence “He talks” the insertion of the verbal affix “–s” relies on information contained in the subject-noun phrase, namely the features PERS(ON) and NUM(BER) and their values PERS=3 and NUM=SG. These features are unified in S as shown in Figure 14.2. In other words, the need to store grammatical information on PERS and NUM during sentence generation illustrates the non-linearity of this morphological process.

In the design of PT, the point of unification is related to a hierarchy of processability that reflects the time course of real time processing as detailed in Levelt (1989). In this way a range of morphological and syntactic processes can be aligned with a universal hierarchy of processability yielding developmental trajectories for the given target languages as shown in several chapters of Pienemann (2005). The hierarchy that results from a comparison of the points of feature unification can be illustrated on the basis of Figure 14.2 which shows that the example structures described in Figure 14.2 can be ordered as follows:

1. (1)   No exchange of grammatical information (= no unification of features),
2. (2)   Exchange of grammatical information within the phrase,
3. (3)   Exchange of grammatical information within the sentence.

Once one applies this hierarchy to ESL morphology, the following developmental trajectory can be predicted:

1. (1)   past –ed
2. (2)   plural –s (det + N)
3. (3)   third-person –s.

In order to appreciate the universal nature of PT, it is crucial to consider that the hierarchy illustrated in Figure 14.2 is not language-specific and that, in principle, it applies to the transfer of grammatical information in any language. The application of the full processability hierarchy to the syntax and morphology of specific languages will, of course, involve more detail of the LFG formalism. It should be noted here that PT does not predict morphology and syntax to develop in tandem. Instead, in L2 acquisition “developmental trailers” (Pienemann, 1998a) may develop which can be brought in line through formal intervention.

A lexically driven grammar (second principle) stores grammatical information in the lexicon. For instance, the lexical entry for “walked” is marked for past tense and it lists the core argument of the verb as “agent.” This lexical information is required in the assembly of the sentence. The lexically driven nature of sentence generation is an integral part of Levelt's approach. LFG also encodes syntactic properties primarily in the lexicon (cf., Schwarze, 2002, pp. 148–149). This makes LFG particularly suitable for the study of dynamic linguistic systems such as developing learner grammars, because LFG affords a formal account of the linguistic dynamics present in developing learner grammars.

The lexically driven nature of sentence generation is supported by a wide range of psycho-linguistic empirical evidence including research on slips of the tongue and on-line experiments (cf., Levelt, 1989) and was demonstrated in experimental work on sentence production by Pickering et al. (2002, p. 127) which shows that “constituent structure is formulated in one stage” and thus supports the architecture of LFG. Pienemann (1998a, b) showed that every level of the PT hierarchy processing procedures can be captured through feature unification in LFG which in turn shares key characteristics with Kempen and Hoenkamp's (1987) procedural account of language generation.

The third PT principle, lexical mapping, is based on Lexical Mapping Theory (LMT) which is a component of LFG (cf., Bresnan, 2001). LFG has three independent and parallel levels of representation as shown in Figure 14.3: a(rgument) structure, f(unctional) structure and c(onstituent) structure. A-structure is universal. The component parts of f-structure and c-structure are also universal, whereas the specific form of the latter two are language-specific.

Lexical mapping refers to the mapping of argument structure onto functional structure. In PT the default mapping principle is “unmarked alignment” which is based on the one-to-one mapping of semantic roles onto grammatical functions. Naturally, mature languages allow for a much wider range of relationships between argument structure and functional structure (including passives, topicalization etc), and these develop step-wise in SLA. Principles of lexical mapping can account for these developmental processes.

In other words, Unmarked Alignment is the initial state of L2 development. It is based on the one-to-one mapping of the three parallel levels of representation onto each other. Unmarked alignment results in canonical word order. For English this is SVO. Unmarked alignment simplifies language processing for the learner who, at this stage, will analyze the first noun phrase as the agent. This way, canonical word order avoids any kind of exchange of grammatical information during language processing.

It follows from the Unmarked Alignment hypothesis that L2 learners will not have access to L2-specific a-structures for predicates. This implies that in cases where L1 and L2 predicates have different a-structures, one can predict that L2 learners will initially have to map arguments canonically onto the LMT hierarchy of core grammatical functions.

Figure 14.3 Unmarked alignment in LFG

Source: Pienemann et al., 2005, p. 230.

The passive is a good example of this mapping process. In the passive the relationship between argument roles and syntactic functions may be altered as can be seen in the supression of argument roles and altered function-assignment. These alterations for passives are illustrated in examples (1)–(4).

Sentences (1) and (3) describe the same event involving two participants. The difference between the two is that in (3) the constituent a dog that is OBJ in (1) is promoted to SUBJ, and the constituent Peter that is SUBJ in (1) is defocused and realized as ADJ. These alterations of the relationship between argument roles and syntactic functions constitute a deviation from Unmarked Alignment.

Similar predictions can also be derived from the relationship between f-structure and c-structure. One set of such predictions is “packaged” in the TOPIC Hypothesis. To account for developmental dynamics in the relationship between f-structure and c-structure Pienemann et al. (2005) propose the TOPIC hypothesis as in (6).

(6) The TOPIC hypothesis
In second language acquisition learners will initially not differentiate between SUBJ and other discourse functions (e.g., TOP). The addition of an XP to a canonical string will trigger a differentiation of TOP and SUBJ which first extends to non-arguments and successively to core arguments thus causing further structural consequences.

The mapping principles involved in the TOPIC Hypothesis and their structural outcomes are summarized in Table 14.2.

Table 14.2 The topic hypothesis

Discourse principle	c- to f- mapping	Structural outcomes
Topicalization of core arguments	TOP = OBJ	The TOP function is assigned to a core argument other than SUBJ.
↑	↑	↑
XP adjunction	TOP = ADJ	Initial constituent is a circumstantial adjunct or a FOCUS WH-word. TOPIC is differentiated from SUBJECT
↑	↑	↑
Canonical Order	SUBJ = default TOP	TOPIC and SUBJECT are not differentiated.

Data and common elicitation measures

Given that PT focuses on natural language production, the main type of data used in PT based research is conversational data – which are subject to the very time constraints in language production that are a crucial element in PT. In addition, reaction-time data have been used since they can potentially tap directly into the language production process.

Initially, conversational data were collected in “linguistic interviews” that followed thematic guidelines. These were conducted as participant observations where the interviewer actively participates in the discourse. It was found that this approach to data elicitation and data collection has a number of drawbacks. The formal drawback of interviews is the poverty of the data they elicit. Even very long interviews have been shown to elicit very small quantities of data that are relevant for a given research question.

Pienemann and Mackey (1993) and Mackey (1992) developed tasks that were designed on the basis of principles derived from task-based language teaching (Long, 1985) for the purpose of eliciting specific L2 grammatical contexts. Figure 14.4 shows the effect of elicitation tasks on the frequency of three grammatical forms in an ESL study with six informants. Even though the interview lasted longer than the two elicitation tasks, the tasks elicited more relevant structures, and they elicited those structures that were targeted (i.e., 3rd-person-s in “Habitual Actions” and questions in “Picture Differences.”) In an extensive study Mackey (1992) demonstrated that these trends can be generalized and that tasks serve to speed up the data elicitation process.

As can be seen in Figure 14.4, elicitation tasks serve the purpose of eliciting relevant data (see also Mackey, 1992). For instance, the study of third-person-s marking requires a large set of contexts for third-person-s. This will allow the researcher to decide if the verbal marker is supplied or not. If no context appears, no conclusion can be drawn. However, even the presence of a number of morphological markers is no guarantee that these are based on productive IL rules. In order to exclude the use of formulae and chunks the researcher needs to check lexical and morphological variation (i.e., the same morpheme on different words and the same word with different morphemes). These descriptive methods are described in more detail in Pienemann (1998a).

Figure 14.4 Frequency of obligatory contexts in tasks and interviews

Source: Pienemann and Mackey, 1993.

The interpretation of corpus data depends on the acquisition criterion that is used. Pienemann (1998a) makes a case for the use of the emergence criterion. The basic point is this: Accuracy criteria (e.g., 80 percent suppliance) are arbitrary. This is illustrated in Figure 14.5 which shows three different routes for the development of the rate of suppliance of grammatical structures. Obviously, the three routes have different gradients. Therefore the order of acquisition of the three structures is c > b > a using a 50 percent criterion and c > a > b using a 100 percent criterion. In other words, accuracy rates do not permit firm conclusions about the state of the IL. In contrast, the point of emergence (of a linguistic form) is not arbitrary. It also permits one to trace development over time, and it operationalizes the first point in time when a structure can be processed. The emergence criterion is operationalized differently for syntax and morphology, taking into account the status of formulae (cf., Pienemann, 1998a). Any test of the PT hierarchy would need to be based on the same definition of the acquisition criterion and on sufficiently large sets of relevant spontaneous production data. Naturally, spontaneous production data cannot be directly compared with other types of data, such as grammatical judgment data because both sets of data tap into different skills and /or types of knowledge.

Figure 14.5 Accuracy and developmental trajectories

We now turn to empirical verification. Since the publication of the first systematic presentation of PT (Pienemann, 1998a) the theory has been tested in many different contexts and has been applied to a range of typologically different L2s. For the purpose of this handbook chapter we will summarize one recent study in some detail and give an overview of some of key research projects that have been carried out within the PT framework.

Our exemplary study focuses on transfer. As mentioned above, PT implies that L1 forms can be transferred to the L2 only when they can be processed in the developing L2 system. Pienemann et al. (2011) tested this hypothesis in a cross-sectional study of Swedish SLA based on seven German ab-initio students who performed unseen communicative tasks eliciting V2 structures after only 30 minutes of exposure to Swedish (on a one-to-one basis). This intensive format of input was used with “square one beginners” to be sure to be tapping into the initial state. The study focuses on V2 because the L1 and the L2 are both V2 languages, and V2 appears at stage 5 in the PT hierarchy. Therefore, a full transfer hypothesis would predict that these learners will transfer V2 from their L1 (German) to L2 Swedish, whereas PT predicts that they won't because they are not ready to process V2 in the L2 system.

Figure 14.6 gives an overview of the results. All learners produced SVO sentences, and five of seven learners also produced the context for V2 in L2 Swedish (here: preposed adverbials). In other words, the intensive input session generated successful L2 learning at the initial state, and most learners “attempted” the V2 structure. However, none of the learners produced even a single instance of V2 in the L2.

Figure 14.6 The effect of 30 minutes’ exposure to L2 Swedish with L1 German

The non-application of V2 cannot be explained by transfer from English as the strongest L2 because three of the informants have a V2 language as their strongest L2. The study also focused on the position of the negator. The learners used proverbal negation but they could not transfer do-insertion from English.

We turn to a discussion of some of the key projects that utilize and support PT. Pienemann and Håkansson (1999) applied the PT hierarchy to Swedish L2 morpho-syntax. Their research is based on an extensive overview of the majority of studies carried out on Swedish SLA over a period of more than two decades. Testing this set of predictions in the large number of studies reviewed by Pienemann and Håkansson produced very solid empirical support for the overall approach. Glahn et al. (2001) tested PT with specific reference to affixation in attributive and predicative adjectives in Scandinavian languages (cf., also Hammarberg, 1996). When intervening variables, such as gender assignment, are “factored out” (cf., Pienemann, 1998a, pp. 159ff.) the data strongly support the predictions made by the theory.

The application of PT to Japanese SLA constitutes a crucial test case for the theory because all other languages PT had initially been applied to (i.e., German, English, and Swedish) are closely related to each other and share a number of typological features. For instance, they are all configurational languages with verb-second constraints. In contrast, Japanese is a non-configurational, agglutinative language. The crux of the typological plausibility test for PT is to demonstrate that the universal architecture of grammatical information flow entailed in PT can make testable and correct predictions for the course of Japanese L2 development. Kawaguchi (2005) demonstrated that the developmental predictions she made for Japanese as L2 on the basis of PT are bourne out by longitudinal and cross-sectional learner data.

Itani-Adams (2003) studied the bilingual first language acquisition of a Japanese-Australian child. Her work utilizes the developmental trajectory of English as a second language established by Pienemann (1998a) and the developmental trajectory of Japanese as a second language established by Di Biase and Kawaguchi (2002) and Kawaguchi (2005). Itani-Adams found that both languages develop following the hierarchy predicted by PT (lexical > phrasal > inter-phrasal). She used PT as a matrix for a comparison of language development across the two first languages of the informant. She found the three different types of morphology (i.e., lexical, phrasal, and interphrasal morphology) did not develop in synchrony in two languages. Iwasaki (2003) studied the acquisition of Japanese morphosyntax by a child a seven-year-old Australian boy in a naturalistic environment and found the same developmental trajectory as Di Biase and Kawaguchi (2002).

Zhang (2001 and 2005) applied PT to modern standard Chinese, another language that is typologically distant from Germanic languages. Modern standard Chinese is an isolating and topic-prominent language. Zhang identified five grammatical morphemes and related them to three different PT levels. Zhang carried out a one-year longitudinal study with five learners using communicative tasks. Her data show that the above-mentioned morphemes are acquired in the sequence predicted by PT. Gao's (2004) study supports these findings.

Di Biase (cf., Di Biase and Kawaguchi, 2002) applied the PT hierarchy to Italian as L2. One important corollary of his research is the conclusion that the mechanics of subject-verb agreement may vary between typologically different languages. This analysis is not only congruent with the data presented by Di Biase and with his implementation of PT into Italian, but also with cross-linguistic on-line studies of agreement in English and Romance languages by Vigliocco et al. (1996) and Vigliocco et al. (1995).

Empirical research carried out in the PT framework yielded precise descriptions of ESL and other developmental trajectories and of learner variation. This research has been used as an objective psycholinguistic basis of measurement in educational evaluation studies. For instance, Keβler (2006) compared levels of attainment in primary and secondary ESL programs in German schools utilizing PT-based descriptions of ESL development. Liebner (2006) studied the acquisition of English as L2 by German and Swedish primary school students in immersive and nonimmersive contexts and compared learner variation in the different contexts demonstrating that non-immersive and non-communicative contexts are more likely to be associated with simplified varieties of learner language. Berti and Di Biase's (2002) and Di Biase's (2002) studies show the effectiveness of L2 teaching based on a syllabus informed by the PT hierarchy in conjunction with form-focused teaching. PT has also been used as a basis for profiling natural and formal L2 learners of Italian as an L2 (Bettoni and Di Biase, 2005). This study confirms the results of Di Biase's (2002) Australian study of Italian L2 development in a formal context.

Håkansson et al. (2003) utlized the cross-linguistic capacity of PT to measure the two languages of bilingual informants on one and the same scale. Håkansson et al. studied the acquisition of Swedish-Arabic bilingual children with and without specific language impairment. Based on the above PT hierarchy for Swedish, Mansouri's (2000) Arabic PT levels and further development of this research by Håkansson and Mansouri, Håkansson et al. (2003) were able to measure language development of bilingual informants using compatible scales for both languages. Key issues arising from the comparison between impaired and non-impaired language development are taken up by Håkansson (2001).

PT-based research is presented at annual international workshops which have generated several volumes on a variety of theoretical and applied topics, including aphasia, bilingual language development, Task-based Language Teaching (TBLT), transfer and testing (Keβler, 2008b; Keβler and Keatinge, 2009; Mansouri, 2007; Bettoni and Di Biase, under review). Work presented at these workshops and other PT-based research appears in a book series published by Benjamins.

Applications

From the outset, PT has been developed with applications to diagnosing language development and language teaching in mind. In this section we will briefly discuss the following practical applications of PT:

•    Diagnosing language development,
•    Developmental approaches to the L2 classroom,
•    Interfaces between TBLT and PT.

Diagnosing language development

The connection between PT and diagnosing language development is obvious. A theory that can account for developmental trajectories and that is based on an abundance of empirical research detailing standard developmental trajectories for a range of L2s (e.g., Johnston, 1997; Keβler, 2006, 2007; Mansouri and Duffy, 2005; Pienemann and Johnston, 1987; Pienemann, 1998a, 2005; Pienemann et al., 2006) is in an excellent position to serve as a point of reference for developmental linguistic profiles.

The Rapid Profile procedure has been developed and refined over a number of years (Keβler, 2006, 2007, 2008a; Keβler and Keatinge, 2008; Mackey et al., 1991; Pienemann, 1990, 1992; Pienemann, 2003; Pienemann et al., 1988; Pienemann et al., 1993; Pienemann and Keβler, 2007;). In this approach a trained analyst identifies a set of 25 developmental features in the speech of a learner. This ocurrs on-line, using profiling software. The speech sample is elicited using communicative tasks as elicitation devices. After about 10–15 minutes a reliable profile of the learner's L2 can be obtained. Other L2s work in a similar manner. At this stage, Rapid Profile has been implemented for the following L2s: English, German, and Turkish.

The software serves the following functions. It records the observations made by the analyst and applies the emergence criterion that has been implemented into an expert system. In addition, it specifies the accuracy of IL structures in the given samples. On this basis it provides on-line feedback to the analyst about the current profile. The software also contains a training environment that can certify an analyst, a database of profiles that can be used in a school context, and a database of tasks that can be used in connection with on-line feedback. The analysis provided by the system also allows the analyst to assess aspects of the lexicon and of variational features as well as grammatical accuracy.

The refinement of Rapid Profile has been driven in part by studies of its reliability. The most recent study is that by Keβler (2006, 2007) who compared the Rapid Profile analysis of 21 samples by newly trained analysts with the the full linguistic analysis of the same samples. Overall, he found an 86 percent agreement on the level of acquisition between the on-line observation and the full analysis. The remaining 14 percent were due to under-observations by one level.

Since Rapid Profile is criterion-referenced, it identifies both what the learner can do as well as what he or she cannot do at a given point in the ESL learning process. The profile produced by this procedure permits the teacher to pinpoint those developmental features that are learnable by a particular student at the given point in time and that can thus be focused on in teaching.

A brief word about linguistic profiling and proficiency rating: As Brindley (1998) pointed out, current rating scales are not based on SLA research and they do not reflect L2 development even if their authors claim the opposite. This does not mean, however, that Rapid Profile was designed to replace rating scales. Such an undertaking would be pointless for the near future, because linguistic profiling and rating scales have rather different objectives. Profiling, by its very nature, aims at maximum precision whereas rating scales aim at maximum scope (Pienemann and Keβler, 2007). At the current state of our knowledge we cannot have both in one procedure. Therefore both are needed until we have a more complete understanding of SLA. Until that time the interface between the two approaches needs to be explored for the benefit of the practitioner (cf., Ferrari and Nuzzo, 2009; Lenzing and Plesser, 2010).

Developmental approaches to the L2 classroom

The Teachabilty Hypothesis (Pienemann, 1984, 1989, 1998a) is the second major theory-practice interface that is part of PT. The Teachability Hypothesis predicts that stages of SLA cannot be skipped through formal intervention and that variational features are not subject to these constraints. The constraints implied in the Teachability Hypothesis have been tested and substantiated in a number of empirical studies which are listed in Table 14.3 (cf., Mansouri and Duffy, 2005 and Keβler et al., 2011 for overviews).

Table 14.3 Studies relating to the Teachability Hypothesis

Study	TL	Learners’ L1	Design	Findings
Pienemann (1984)	German	Italian	Pre-test, Post-test control group design: whether stages can be skipped	Stages of acquisition cannot be skipped
Ellis (1989)	German	English	Pre-test, Post-test control group design: formal versus naturalistic instruction	Support for the Teachability Hypothesis
Boss (1996)	German	English/Chinese	Oral language production compared to taught syllabus as opposed to PT sequence	Learners progressed in the predicted order regardless of the taught syllabus
Spada and Lightbown (1999)	ESL	French	Pre-test, Post-test control group design: whether stage skipping is possible	Inconclusive: no support for Teachability Hypothesis1
Dyson (1996)	ESL	Spanish	Longitudinal study of ESL development with a syllabus based on teachable forms	Overall support for the Teachability Hypothesis despite inter-learner variation
Mansouri and Duffy (2005)	ESL	Chinese/Korean/ Thai	Pre-test, Post-test control group design: developmental versus reversed order group	Support for the Teachability Hypothesis

Note:

1 This study was inconclusive because the informants had already acquired the test structure in the pre-test. Also, the studyt is based on the false assumption that the Teachability Hypothesis predicts that “timed intervention” will promote acquisition.

Whereas the Teachability Hypothesis delineates constraints on teachability, “timed intervention” (Long, 1988) goes one step further by assuming that formal intervention will promote acquisition if it occurs at a time when the learner is developmentally ready for the taught form. Di Biase (2002, 2008) demonstrated that learners who received developmentally timed form-focused instruction (Long, 1991) traversed the developmental hierarchy more quickly and more successfully than learners without timed (and form-focused) instuction.

From a PT perspective, timed intervention needs to be driven by the developmental readiness of the learners. This approach is facilitated by assigning a developmental profile to every student using Rapid Profile. In this way, formal intevention can focus on developmentally relevant forms for every individual. As Pienemann (1985) pointed out, this approach does not imply the inclusion of non-target IL forms in the syllabus.

Interfaces between task-based language teaching and PT

Lightbown (1998, p. 179) raised the concern that the heterogeneity of learners’ profiles in class “. . . would make developmentally targeted teaching very difficult to organize.” Keβler (2008a) shows that the heterogeneity of linguistic profiles does not preclude developmentally timed intervention if it occurs in task-based language teaching (Long, 1985). Several studies (e.g., Keβler, 2008a; Mackey, 1992) have shown that learners at different levels of acquisition use different linguistic forms in the same task—nevertheless successfully communicating with each other. Therefore one syllabus can be used in heterogeneous classes enabling the teacher to focus on different forms for learners with different profiles while the negotiation of meaning is coherent for the whole group. In this context, the more advanced learner can also provide scaffolds for his/ her partner while working on the same task (Keβler, 2008a), thus priming (McDonough and Mackey, 2008) the less developed learner.

Keβler (2008a) proposes a “diagnostic task cycle” for the implementation of developmentally timed instruction in TBLT. The cycle starts with the task-based elicitation of interlanguage samples for linguistic profiling. This first step generates the diagnostic basis that informs the design of tasks in the cycle. Once the teacher has ascertained the learners’ developmental profiles, the task-as-workplan (cf., Seedhouse, 2005) phase begins with pre-task activities to provide the necessary preconditions for the learning task and interactive language learning to be successful – such as knowledge of the lexical fields relevant for the learning task. Once this preparatory work has been completed, the class proceeds to the actual task work which is now developmentally appropriate and communicatively manageable.

When the task-as-workplan-phase is completed the focus of instruction shifts from timed intervention to more conventional communicative activities that focus on the negotiation of meaning in order to use the newly acquired linguistic and communicative competence in new and different contexts. Given the developmental timing of the instruction, this phase also provides an opportunity for the teacher to give corrective feedback on learnable forms.

The diagnostic task cycle is completed by a new round of profiling. The second profiling exercise serves as an evaluation of language development and, at the same time, as a diagnostic basis for the next cycle of task-based activities.

Future directions

One key area that requires attention in the further development of PT is the interface between the learner's developmental grammar and its effect on discourse and interaction (cf., Mackey, Abbuhl and Gass, Chapter 1, this volume). The basic assumption is that the constraints on L2 grammar entailed in PT also constrain L2 discourse and interaction. For instance, the Unmarked Alignment Hypothesis and the TOPIC Hypothesis imply that at the initial state of L2 acquisition learners have no choice other than to produce sentences using canonical word order (SVO for English). This means that at the initial state learners lack the grammatical means to express information that is structured in a non-linear way. For instance, at this stage learners do not have the freedom to structure information according to “given” and “new” information. Instead, the agent-subject always needs to appear as the first constituent.

A further focus of future research on PT is on automatization. For instance, Kawaguchi and Di Biase (in press) carried out a number of experiments to study how L2 performance in the formation of passives improves due to automatization. Pienemann (2006) studied the relationship between level of acquisition and word access utilizing Segalowitz’ (2003) operationalization of automatization (cf., Segalowitz and Trofimovich, Chapter 11, this volume). The purpose of this research is to gain a better understanding of the learner's development from the point of emergence to the mastery of linguistic forms.

In the area of diagnosing L2 development, software has been developed and trialled that permits a fully automatic analysis of written interlanguage samples. The software is based on the systematicity of L2 development and regularities in time-constrained L2 writing. This approach may serve to make profiling more accessible to practitioners.

As we pointed out above, PT derives hypotheses about the L2 initial state (and thus about the “logical problem”) from constraints on processing, lexical mapping and the architecture of LFG. One example is L2 initial word order. This line of thought has great potential for further investigation which can be based on a systematic integration of the processing and mapping phenomena covered by PT as well as on the integration of universal principles inherent in LFG.

References

Andersen, R. (1984). The one-to-one principle. Language Learning, 34, 77–95.

Berti, P. and Di Biase., B. (2002). Form-focused instruction and the acquisition of nominal morphology in primary school learners of Italian L2. Italiano e Scuola NSW Italian Teachers Bulletin, 7(1), 10–14.

Bettoni, C. and Di Biase, B. (2005). Sviluppo obbligato e progresso morfosintattico – un caso di Processabilità in Italiano L2. Italiano Lingua Seconda, 3(7), 27–48.

Bettoni B. and Di Biase B. (Eds.) (under review). Processability theory: Current issues in theory and application. Amsterdam: Benjamins.

Boss, B. (1996). German grammar for beginners – the teachability hypothesis and its relevance to the classroom. In C. Arbones Sola, J. Rolin-Ianziti, and R. Sussx (Eds.), Proceedings of the conference on “Who's afraid of teaching grammar?” Working papers in applied linguistics (Vol. 1, pp. 93–103). Brisbane: Centre for Language Teaching and Research, University of Queensland, Australia.

Bresnan, J. (2001). Lexical-functional syntax. Malden, MA: Blackwell.

Brindley, G. (1998). Describing language development? Rating Scales and SLA. In L. F. Bachman and A. D. Cohen (Eds.), Interfaces between second language acquisition and language testing research (pp. 112–140). Cambridge: Cambridge University Press.

Clahsen, H. (1984). The acquisition of German word order: A test case for cognitive approaches to L2 development. In R. Anderson (Ed.), Second Languages (pp. 219–242). Rowley, MA: Newbury House.

Clahsen, H. (1992). Learnability theory and the problem of development in language acquisition. In J. Weissenborn, H. Goodluck, and T. Roeper (Eds.), Theoretical issues in language acquisition: Continuity and change (pp. 53–76). Hillsdale, NJ: Lawrence Erlbaum.

Di Biase, B. (2002). Focusing strategies in second language development: A classroom-based study of Italian L2 in primary school. In B. Di Biase (Ed.), Developing a second language: Acquisition, processing and pedagogy issues in Arabic, Chinese, English, Italian, Japanese, Swedish (pp. 95–120). Melbourne: Language Australia.

Di Biase, B. (2008). Focus-on-form and development in L2 learning. In J. U. Keβler (Ed.), Processability approaches to second language development and second language learning (pp. 197–219). Newcastle upon Tyne, UK: Cambridge Scholars Publishing.

Di Biase, B. and Kawaguchi, S. (2002). Exploring the typological plausibility of Processability Theory: Language development in Italian second language and Japanese second language. Second Language Research, 18(3), 274–302.

Dyson, B. (1996). The debate on form-focused instruction: A teacher's perspective. Australian Review of Applied Linguistics, 19(2), 59–78.

Ellis, R. (1989). Are classroom and naturalistic acquisition the same? A study of the classroom acquisition of German word order rules. Studies in Second Language Acquisition, 11, 303–328.

Ferrari, S. and Nuzzo, E. (2009, July). Exploring interfaces between PT and other SLA related measures. Paper presented at the 9th International Symposium on Processability Theory, Bilingualism and Second Language Acquisition, Ludwigsburg, Germany.

Gao, X. (2004). Noun phrase morphemes and topic development in L2 Mandarin Chinese: A processability perspective. Unpublished doctoral dissertation. Victoria University of Wellington, New Zealand.

Glahn, E., Håkansson, G., Hammarberg, B., Holmen, A., Hvenekilde, A., and Lund, K. (2001). Processability in Scandinavian second language acquisition. Studies in Second Language Acquisition, 23, 389–416.

Håkansson, G. (2001). Tense morphology and verb-second in Swedish L1 children, L2 children and children with SLI. Bilingualism: Language and Cognition, 4, 85–99.

Håkansson, G., Salameh, E. -K., and Nettelbladt, U. (2003). Measuring language development in bilingual children: Swedish-Arabic children with and without language impairment. Linguistics, 41, 255–288.

Hammarberg, B. (1996). Examining the processability theory: The case of adjective agreement in L2 Swedish. In E. Kellerman, B. Welters, and T. Bongaerts (Eds.), Toegepaste taalwetenschap in artikelen 55: EUROSLA 6. A Selection of Papers, Nijmegen 1996 (pp. 75–88).

Itani-Adams, Y. (2003, September). From word to phrase in Japanese-English bilingual first language acquisition. Paper presented at the MARCS seminar, University of Western Sydney, Sydney, Australia.

Iwasaki, J. (2003, July). The acquisition of verbal morphosyntax in JSL by a child learner. Paper presented at 13th Biennial Conference of the JSAA, Brisbane, Australia.

Johnston, M. (1997). Development and variation in learner language. Unpublished doctoral dissertation. Australian National University, Canberra, Australia.

Kawaguchi, S. (2005). Argument structure and syntactic development in Japanese as a second language. In M. Pienemann (Ed.), Cross-linguistic aspects of processability theory (pp. 253–298). Amsterdam: John Benjamins.

Kawaguchi, S. and Di Biase, B. (in press). Emergence and automatization in second language development. In C. Bettoni and B. Di Biase (Eds.), Processability theory: Current issues in theory and application. Amsterdam: Benjamins.

Kempen, G. and Hoenkamp, E. (1987). An incremental procedural grammar for sentence formulation. Cognitive Science, 11, 201–258.

Keβler, J. -U. (2006). Englischerwerb im Anfangsunterricht diagnostizieren. Linguistische Profilanalysen und der Übergang von der Primar- in die Sekundarstufe I. Giessener Beiträge zur Fremdsprachendidaktik. Tübingen, Germany: Narr.

Keβler, J. -U. (2007). Assessing EFL-development online: A feasibility study of rapid profile. In F. Mansouri (Ed.), Second language acquisition research: Theory-construction and testing (pp. 119–143). Newcastle upon Tyne, UK: Cambridge Scholars Press.

Keβler, J. -U. (2008a). Communicative tasks and second language profiling: Linguistic and pedagogical implications. In J. Eckerth and S. Siepmann (Eds.), Research on task-based language learning and teaching: Theoretical, methodological and pedagogical perspectives (pp. 291–310). New York: Peter Lang.

Keβler, J. -U. (Ed. ) (2008b). Processability approaches to second language development and second language learning. Newcastle upon Tyne, UK: Cambridge Scholars Publishing.

Keβler, J. -U. and Keatinge, D. (2008). Profiling oral second language development. In J. -U. Keβler (Ed.), Processability approaches to second language development and second language learning (pp. 167–197). Newcastle upon Tyne, UK: Cambridge Scholars Publishing.

Keβler, J. -U. and Keatinge, D. (Eds. ) (2009). Research in second language acquisition. Evidence across languages. Newcastle upon Tyne, UK: Cambridge Scholars Publishing.

Keβler, J. -U., Liebner, M., and Mansouri, F. (2011). “Teaching”. In M. Pienemann and J. -U. Keβler (Eds.), Studying processability theory. An introductory textbook (pp. 149–156). Amsterdam: John Benjamins.

Larsen-Freeman, D. and Long, M. H. (1991). An introduction to second language acquisition research. London: Longman.

Lenzing, A. and Plesser, A. (2010). “Challenging the Scope-Presission Dilemma in language testing. Paper presented at the 10th International Symposium on Processability Theory, Bilingualism and Second Language Acquisition, Sydney.

Levelt, W. J. M. (1989). Speaking: From intention to articulation. Cambridge, MA: MIT Press.

Liebner, M. (2006). Variationsverhalten in verschiedenen Lernergruppen. In M. Pienemann, J. -U. Keβler, and E. Roos (Eds.), Englischerwerb in der Grundschule. Ein Studien- und Arbeitsbuch. Paderborn, Germany: Schöningh/UTB.

Lightbown, P. M. (1998). The importance of timing in focus on form. In C. Doughty and J. Williams (Eds.), Focus on form in second language acquisition (pp. 177–196). Cambridge: Cambridge University Press.

Long, M. H. (1985). A role for instruction in second language acquisition: Task-based language teaching. In K. Hyltenstam and M. Pienemann (Eds.), Modelling and assessing second language acquisition (pp. 77–99). Clevedon, UK: Multilingual Matters.

Long, M. H. (1988). Instructed interlanguage development. In L. Beebe (Ed.), Issues in second language acquisition: Multiple perspectives (pp. 115–141). Cambridge, MA: Newbury House.

Long, M. H. (1991). Focus on Form: A design feature in language teaching methodology. In K. de Bot, R. Ginsberg, and C. Kramsch (Eds.), Foreign language research in cross-cultural perspective (pp. 39–52). Amsterdam: John Benjamins.

Mackey, A. (1992). Targeting morpho-syntax in children's ESL: An empirical study of the use of interactive goal-based tasks. Working Papers in Educational Linguistics, 10(1), 67–91.

Mackey, A., Pienemann, M., and Thornton, I. (1991). Rapid PROFILE: A second language screening procedure. Language and Language Education: Working Papers of the National Languages Institute of Australia, 1, 61–82.

Mansouri, F. (2000). Grammatical markedness and information processing in the acquisition of Arabic as a second language (pp. 1–260). Munich, Germany: LINCOM EUROPA Academic Publishers.

Mansouri, F. (2007). Second language acquisition research: Theory-construction and testing. Newcastle upon Tyne, UK: Cambridge Scholars Press.

Mansouri, F. and Duffy, L. (2005). The pedagogic effectiveness of developmental readiness in ESL grammer instruction. Australian Review of Applied Linguistics, 28(1), 81–99.

McDonough, K. and Mackey, A. (2008). Syntactic priming and ESL question development. Studies in Second Language Acquisition, 30(1), 31–47.

Meisel, J. M. (1983). Strategies of second language acquisition: More than one kind of simplification. In R. W. Andersen (Ed.), Pidginisation and creolisation as language acquisition (pp. 120–157). Rowley, MA: Newbury House.

Meisel, J. M., Clahsen, H., and Pienemann, M. (1981). Determining developmental stages in natural second language acquisition. Studies in Second Language Acquisition, 3, 109–135.

Pickering, M. J., Branigan, H. P., and McLean, J. F. (2002). Constituent structure is formulated in one stage. Journal of Memory and Language, 46, 586–605.

Pienemann, M. (1984). Psychological constraints on the teachability of languages. Studies in Second Language Acquisition, 6(2), 186–214.

Pienemann, M. (1985). Learnability and syllabus construction. In K. Hyltenstam and M. Pienemann (Eds.), Modelling and assessing second language acquisition (pp. 23–75). Clevedon, UK: Multilingual Matters.

Pienemann, M. (1989). Is language teachable? Psycholinguistic experiments and hypotheses. Applied Linguistics, 10(1), 52–79.

Pienemann, M. (1990). LARC Research Projects 1990. Sydney: NLIA/Language Acquisition Research Center.

Pienemann, M. (1992). Assessing second language acquisition through rapid profile. ms. Sydney.

Pienemann, M. (1998a). Language processing and second language development - processability theory. Amsterdam: John Benjamins.

Pienemann, M. (1998b). Developmental dynamics in L1 an L2 acquisition: Processability theory and generative entrenchment. Bilingualism: Language and Cognition, 1(1), 1–20.

Pienemann, M. (2003). Language processing capacity. In C. Doughty and M. Long (Eds.), Handbook of second language acquisition theory and research (pp. 679–714). Malden, MA: Blackwell.

Pienemann, M. (Ed.) (2005). Cross-linguistic aspects of processability theory. Amsterdam: John Benjamins.

Pienemann, M. (2006). “Applying SLA research to the measurement of bilingual proficiency”. Paper read at the Conference on “Lexical organisation in bilinguals”. University of the West of England, Bristol, 6 January, 2006.

Pienemann, M., Di Biase, B., and Kawaguchi, S. (2005). Extending processability theory. In M. Pienemann (Ed.), Cross-linguistic aspects of processability theory (pp. 199–251). Amsterdam: John Benjamins.

Pienemann, M., Di Biase, B., Kawaguchi, S., and Håkansson, G. (2005). Processing constraints on L1 transfer. In J. F. Kroll and A. M. B. DeGroot (Eds.), Handbook of bilingualism: Psycholinguistic approaches (pp. 128–153). New York: Oxford University Press.

Pienemann, M. and Håkansson, G. (1999). A unified approach towards the development of Swedish as L2: A processability account. Studies in Second Language Acquisition, 21(3), 383–420.

Pienemann, M. and Johnston, M. (1987). Factors influencing the development of language proficiency. In D.Nunan (Ed.), Applying second language acquisition research (pp. 45–141). Adelaide, Australia: National Curriculum Resource Centre.

Pienemann, M., Johnston, M., and Brindley, G. (1988). Constructing an acquisition-based procedure for second language assessment. Studies in Second Language Acquisition, 10(2), 217–243.

Pienemann, M., Johnston, M., and Meisel, J. (1993). The multidimensional model, linguistic profiling, and related issues. A reply to Hudson. Studies in Second Language Acquisition, 15(4), 495–503.

Pienemann, M. and Keβler, J.-U. (2007). Measuring bilingualism. In P. Auer and L. Wei (Eds.), Handbook of applied linguistics: Vol. 5. Multilingualism and multilingual communication (pp. 247–275). Berlin/New York: Mouton de Gruyter.

Pienemann, M., Keβler, J.-U. (eds.) (2011). Studying processability theory. An introductory textbook. Amsterdam: John Benjamins.

Pienemann, M., Keβler, J. -U., and Itami-Adams, Y. (2011). Comparing levels of processability across languages [Special issue]. International Journal of Bilingualism, 128–146.

Pienemann, M., Keβler, J. -U., and Lenzing, A. (forthcoming). Transfer at the initial state. In J.-U. Keβler and M. Liebner (Eds.), Developing and Assessing Second Languages. Amsterdam: John Benjamins.

Pienemann, M., Keβler, J. -U., and Liebner, M. (2006). Englischerwerb in der Grundschule: Untersuchungsergebnisse im Überblick. In M. Pienemann, J.-U. Keβler, and E. Roos (Eds.), Englischerwerb in der Grundschule. Ein Studien- und Arbeitsbuch (pp. 67–88). Paderborn, Germany: Schöningh/UTB.

Pienemann, M. and Mackey, A. (1993). An empirical study of children's ESL development and Rapid Profile. In P. McKay (Ed.), ESL development: Language and literacy in Schools. Vol. 2 (pp. 115–259). Canberra: National Languages and Literacy Institute of Australia.

Schwarze, C. (2002). Representation and variation: On the development of Romance auxiliary syntax. In M. Butt and T. Holloway King (Eds.), Time over matter: Diachronic perspectives in morphosyntax (pp. 105–141). Stanford, California: CSLI Publications.

Seedhouse, P. (2005). “Task” as research construct. Language Learning, 58(3), 533–570.

Segalowitz, N. (2003). Automaticity and second languages. In C. Doughty and M. Long (Eds.), Handbook of second language acquisition theory and research (pp. 382–408). Malden, MA: Blackwell.

Slobin, D. I. (1973). Cognitive prerequisites for the development of grammar. In C. A. Ferguson and D. I. Slobin (Eds.), Studies of child language development (pp. 175–208). New York: Holt, Rinehart and Winston.

Spada, N. and Lightbown, P. M. (1999). Instruction, first language influence and developmental readiness in second language acquisition. The Modern Language Journal, 83(1), 1–22.

Vigliocco, G., Butterworth, B., and Garrett, M. F. (1996). Subject-verb agreement in Spanish and English: Differences in the role of conceptual constraints. Cognition, 61, 261–298.

Vigliocco, G., Butterworth, B., and Semenza, C. (1995). Constructing subject-verb agreement in speech: The role of semantic and morphological factors. Journal of Memory and Language, 34, 186–215.

Wexler, K. (1982). A principle theory for language acquisition. In E. Wanner and L. R. Gleitman (Eds.), Language acquisition: The state of the art (pp. 288–315). Cambridge: Cambridge University Press.

Wexler, K. and Culicover, P. (1980). Formal principles of language acquisition. Cambridge, MA: MIT Press.

White, L. (1991). Second language competence versus second language performance: UG or processing strategies? In L. Eubank (Ed.), Point counterpoint. Universal grammar in the second language (pp. 167–189). Amsterdam: John Benjamins.

Zhang, Y. (2001). Second language acquisition of Chinese grammatical morphemes: A processability perspective. Unpublished doctoral dissertation. The Australian National University, Canberra, Australia.

Zhang, Y. (2005). Processing and formal instruction in the L2 acquisition of five Chinese grammatical morphemes. In M. Pienemann (Ed.), Cross-linguistic aspects of processability theory (pp. 155–177). Amsterdam: John Benjamins.