Following Gass & Selinker (2008, p. 5), I understand the area of second language acquisition to be a complex one whose focus is the “attempt to understand the processes underlying the learning of a second [or foreign] language” (see Gass & Selinker, 2008, p. 5; Norris & Ortega, 2000, p. 717) after the first has been learned.17 The study of second language acquisition presents “a firm basis for [language teaching] methodologies in language learning” (Gass & Selinker, 2008, p. 3).
The scope of most grammars shows that the acquisition of a foreign language entails more than just grammar (e.g., sound systems, syntax, morphology and the lexicon, semantics and pragmatics; Gass & Selinker, 2008). In contrast, the CING only contains written/textual information on grammar, supported by an English-German language corpus. In the following discussion of the theoretical background to learning English grammar with the CING, I will thus focus on those theories of second language acquisition (SLA) applicable to the learning situation with the program.18 While the research area of SLA “draws on multidisciplinary
17 “Acquisition” is treated here as similar in meaning to “learning.” While learning entails more than mere
learner improvement in grammar test results, the scope of this study restricts the term’s meaning to the measurable learner improvement documented below.
theoretical and empirical perspectives to address the specific issue of how people acquire a second language” (Larsen-Freeman, 2000, p. 165), it also involves such core topics in cognitive psychology and linguistics as “implicit and explicit learning, incidental and intentional learning, automaticity, attention and memory, individual differences, variation, language processing etc.” (Doughty, Williams & Saunders 2004, p. 6). The majority of research on language acquisition is based on different types of interaction (e.g., one-way information exchange [Long, 1980a] or introspective techniques, e.g., diary study [Brown, 1985], [Ellis 1994]). This is of considerable importance to the following discussion, as the CING is a self-instruction tool with a hypertext structure that presents learners with a quite different material structure and requires different learning strategies than do paper/pencil- based language support tools. Learning with the CING is dependent on the “input text” it provides rather than “input discourse” (c.f. Ellis, 1994), which is a significant factor in the majority of SLA studies.
These SLA theories, then, as represented by the works referred to above, cannot be expected to fully apply to learning with the CING. Once the CING is integrated into a learning situation comparable to those studied by the authors above (e.g., classroom English- language learning, teacher-guided small-group acquisition) different theories will apply than in the self-instruction situation this study considers. The following discussion of the theoretical background always keeps the autonomous learning situation of the CING in mind.
3.2.1 Application of SLA theory in CING research
In the following I will review former research findings on topics relevant to this study. Based on this review these topics were formulated into a list of assumptions I had about learning with the CING, hypertext and autonomous learning. The background to these assumptions, that I also call research hypotheses, will be discussed in the following.
− The more English-language skills learners have, the better they will understand the CING and the better their CING work experience will be (Heller, 2004b).
learning style and strategies (Gass & Selinker, 2008, p.445). While I agree with Gass & Selinker that these factors must be considered when investigating learning holistically, this study’s main aim is to evaluate the CING, and consequently my discussion of the various factors impacting language learning will focus mainly on the unique nature of the program as an autonomous and hypertext self-learning tool. For an overview of other theories on language acquisition and learning see McLaughlin (1987) or Ellis (1985).
− Authentic language material can contain vocabulary that is unfamiliar to intermediate learners. This also applies to the CING grammar material, which is largely based on an earlier CING version written for advanced learners and EFL teachers (Heller, 2004a).19
− Comprehensible language material (rules and language examples) can support language acquisition with the CING. (Long, 1983, Larsen-Freeman & Long, 1991, Ellis, 1994)
− The explicit and formal teaching of language structures supports the short-term acquisition of these structures (Lightbown et al., 1980; Ellis, 1994).
− The grammar structures in the CING material have been sufficiently highlighted for learners to notice them (Doughty, 1991) but might not be sufficient for successful learning (Doughty, 2004; Jourdenais, 1998).
− Feedback promotes learning as it helps learners to notice mistakes in their language production and informs them about their level of applied grammatical competence (Schimmel, 1983; Gass & Selinker, 2008; Bangert-Downs, Kulik, Kulik & Morgan, 1991).
− Learning with a hypertext differs considerably from traditional learning (see Tergan et al., 2000; Tergan & Lechner, 2000; Jacobs, 2004). Successful goal-oriented navigation and orientation in the hypertext structure can require support measures (Brunstein, Naumann & Krems, 2004, p. 3831; Naumann, Waniek, Brunstein & Krems, 2003) the CING does not contain (e.g. explicit scaffolding support: Rouet, 1992; Jonassen, 1993; Jacobsen et al., 1995; guided tours, hypertrails, dynamic paths: Gerdes, 1997).
− Autonomous learning strategies help learners to define learning goals, select relevant learning materials, and apply appropriate learning steps and strategies for their task (Holec, 1981; Little, 1991; Ziegler et al., 2003; Tergan, 2003).
− Learning strategies can be learned (Bannert, 2003).
Preliminary small-scale studies on the CING I conducted with a subject group similar to the one in this study (Heller, 2004a/b) showed that the tool’s language material can cause comprehension problems to learners. Language examples and grammar rules material in the CING was designed on the basis of material developed for advanced learners or teachers. Although other grammar provide even more in-depth analyses of the use of grammar structures, the rule material and language examples in the CING were designed for advanced learners or teachers. This is particularly true for the content materials on the simple past and present perfect.20 Both preliminary studies suggested that learners had problems with unknown vocabulary in the rule and language material. There was also evidence that learners with a better score in their last English-language exam appeared to have fewer difficulties
19 This is more a statement of fact than a hypothesis, but since it stayed relevant throughout the
development of the research measures I include it here. 20
with material comprehension (Heller, 2004b). Disorientation while using the CING’s grammar material and during material search, however, was an issue for all subjects in the study (Heller, 2004a). While disorientation is an issue directly related to learning with hypertext (see below), hypotheses 1 and 2 were formulated to investigate the reasons for learners’ difficulties with the CING material.
Hypothesis 3 calls for only comprehensible input in order to make successful learning possible, pointing to the relevance of input theory. Hypothesis 4 on the merit of instruction in SLA, relates to the domain of “instructed SLA,” including formal instruction theory in general, the teachability hypothesis (Pienemann, 1984, 1985, 1989), and the weak interface hypothesis (Ellis, 1990). I will also investigate the concept of salience of input material (hypothesis 5) and feedback in language learning (during instruction) (hypothesis 6) in SLA studies and relate it to the CING pedagogical grammar content.
3.2.2 Input in SLA
The term “input” in the hypotheses above refers to language and grammar material, including the grammar structures that the CING provides to the user. Ellis writes that “input may come in written or spoken form” (1994, p. 26). The CING, however, is a self-instruction tool, while a lot of SLA theories were built on studies or observations that involved spoken input in learning situations (e.g., classroom studies) but not autonomous learning. Overall in SLA research, input is closely connected to output and interaction (Gass, 2003), but learning with the CING only provides restricted opportunities for the latter. The CING’s aim is rather to help learners hone their receptive skills by studying the authentic language material and practice their productive skills in the grammar exercises.