How to teach grammar in EFL classes. Explicit versus implicit and deductive versus inductive teaching

Table of Contents

Introduction

1. Defining Explicit and Implicit Teaching

2. The Interface Hypotheses

3. Form-focused instruction vs meaning-focused grammar instruction

4. Explicit Instruction

5. The Target Structure

6. Deductive vs. Inductive Grammar Teaching

7. PPP: Presentation, Practice and Production

8. Empirical studies on Deductive and Inductive Teaching

9. Sample lessons
9.1 Introduction and Presentation
9.2 Practice Stage
9.3 Consolidation stage
9.4 Production stage
9.5 Inductive lesson

10. Conclusion

References

Appendices

Introduction

Over the past 20 years and more, a great deal of interest has been paid to grammar teaching in EFL (English as a Foreign Language) and ESL (English as a Second Language) classes. Much of the research has been into which methods and approaches offer the learners the greatest opportunities to learn the target language. However, according to Motha, "[O] pinions on the right approach to learning a language differ as widely as the languages themselves" (Motha, 2013: 3). This helps to describe the discussion that this work covers. The aim of this work is to examine the areas of explicit versus implicit teaching and deductive and inductive teaching, with a focus on teaching grammar in EFL classes.

In Section 1, the dichotomy between explicit and implicit teaching is analyzed by investigating the effectiveness of either approach. Section 2 covers an existing interface between explicit and implicit knowledge and presents the three dominant concepts. Section 3 continues with a comparison between grammar teaching approaches that includes a "focus on form" (Ellis, 2002: 225) and a "focus on meaning" (ibid.) with the overall aim of connecting these two distinct terms in order to discuss the theory and related practice for language teachers to refer to while planning their lessons.

Section 4 presents empirical evidence from two meta-studies by Norris and Ortega (2000) and Spada and Tomita (2010), arguing in favor of explicit over implicit instruction. When considering how grammar is to be taught both efficiently and effectively, the underlying target structure and its theoretical characteristics need to be taken into account in determining whether or not this specific structure is a suitable to be either explicitly or implicitly taught. Section 5 discusses several aspects that are related to the target structure. Grammar can be taught, for example, through deductive or inductive teaching methods. Therefore, these two concepts are introduced in Section 6 and 7 as is the PPP- model as an example of deductive teaching (and which is further discussed in Section 9). Section 8 analyzes empirical data of several studies which have investigated the effectiveness of deductive and inductive grammar teaching. The results, as is demonstrated, are contradictory because of differences in terms of measurement, underlying target structure and/or the participants.

Section 9 presents sample lessons designed for the purpose of this work which consider theoretically valid aspects of how appropriate and effective grammar teaching can be. These lessons also show how the theory can feasibly be linked to the practice of deductive and inductive grammar teaching in a foreign language classroom.

Finally, the conclusion gives a final evaluation of these learner-centered methods before suggesting areas for further research.

1. Defining Explicit and Implicit Teaching

The research area dealing with instructional treatments in Second Language Acquisition (SLA) is characterized by the dichotomy of two umbrella terms for the actual definition of instruction. According to Norris and Ortega the term "explicit instruction" (2000: 437) means that "rule explanation comprised part of the instruction... or if learners were directly asked to attend to particular forms and try to arrive at metalinguistic generalizations on their own" (ibid.). On the other hand, the term "implicit instruction refers to treatments without rule presentations or directions to attend to particular target forms" (ibid.).

The history of research in what type of instructional procedure is the most beneficial for learners of a foreign or second language (FLL) has produced different and also contradicting views. Generally speaking, the aims of instruction in second language learning/ acquisition consider the "internalization of new L2 features and processes of modification and consolidation of L2 knowledge." (De Graaff and Housen 2009: 731).

De Graaff and Housen refer to Krashen's theory that instructional treatment will not increase or accelerate the acquisition of a second language because instruction only enhances the formation of "explicit knowledge" (2009: 732), whereas the "implicit knowledge" (ibid.) is not affected by any sort of instruction (ibid.). Second language acquisition functions rather the same way as first language acquisition, through primary exposure to the target language (ibid.). Therefore, instruction on grammar rules does not show great effects regarding communication and language production abilities (Ellis, 2006: 85).

In order to focus more deeply on the dichotomy between these two concepts of knowledge, two further terms need to be explained, that is, the terms "incidental learning: learning without intention" (Long and Robinson, 2013: 18) and "implicit learning: learning without awareness of what is being learned" (DeKeyser, 2011: 314). Implicit learning, in contrast to incidental learning, involves an underlying objective to learn information and/or a concept, but learners might learn or acquire new knowledge which is not the focus of the planned lesson (ibid.).

However, the underlying question of how second language acquisition is affected through either learning approach has not yet been clearly answered. Rosa and O'Neill question whether incidental learning without consciously paying attention to target structures of the second language is even possible (1999: 512) and, if this is the case, the results of learning remain somewhat insignificant (ibid.: 513). They further differentiate between input and "intake" (ibid.: 531) which can be characterized as the limited amount of the presented target language input that learners actually focus on while they are processing the input.

This definition of intake implies a gap between input and intake, that is, individual differences influence the amount of intake and therefore the amount of potential learning from the input (ibid.). This observation is supported by theory concerning input processing: learners at the beginning of their acquisition processes are not yet able to notice certain forms from the input they receive because they primarily intend to understand the message that is being conveyed (Loewen, 2015: 85). Such difficulties can be overcome if structures that should be implicitly learned by the learner being exposed to the target language also focus on "form- meaning and function relationships" (ibid.).

However, the implied criticism of implicit learning through implicit instruction is counterbalanced by successful and proficient acquisition of a second language through "immersion programs" (Long and Robinson, 2013: 20). The fluent speech of learners clearly and significantly increases through being provided with massive input but, on the other hand, these learners tend to have significant gaps in terms of accurate production when compared to native speakers of that language (ibid, 21). Empirical evidence for implicit learning can be found in learners' abilities to detect "form-meaning connections or phonological rules from the input" (Leow, 2015: 193). Learners will be able to process some input in terms of vocabulary, for example, but more complex and cognitively demanding structures will not be noticed from the input only (DeKeyser, 2013: 43).

In addition, the specifics of the target structure have wider implications on learners' potential acquisition. Krashen' s view of how language is acquired is that learners do not consciously learn new information and also what they explicitly learn is not related to increasing proficiency in spontaneous speech production (Ellis, N.C., 2007: 19). Furthermore, explicitly teaching structures which tend to be easier to comprehend should be focused upon, rather than teaching more complex structures in this way because they tend to be barely learnable (Ellis, 2002: 225; Ellis, 2006: 87).

The relationship between the two types of knowledge are discussed in more detail in the following because the different kinds of knowledge (i.e. implicit and explicit) need to be explained in order to clarify exactly what these terminologies mean. For this thesis, Ellis provides the framework for defining both terms:

Implicit knowledge is procedural, is held unconsciously and can only be verbalized if it is made explicit. It is accessed rapidly and easily and thus available for use in rapid, fluent communication (2005c: 36).

This definition of implicit knowledge contains the overall objective of second language teaching, that is, the learners eventually have the ability to communicate and therefore, at some point, develop this type of knowledge (ibid.).

On the other hand, explicit knowledge is defined as:

the declarative and often anomalous knowledge of the phonological, lexical, grammatical, pragmatic and socio-critical features of an L2 together with the metalanguage for labelling this knowledge. (Ellis 2004; cited in Ellis, 2005c: 36)

Learners can access their explicit knowledge through conscious and active procedural cognitive processes and also use it as a resource if they notice certain problems in their existing interlanguage, their second language abilities (ibid.).

Distinguishing between implicit and explicit knowledge in terms of general specifics is, however, insufficient in order to draw clear distinctions of these two terms. It is therefore necessary to consider how implicit and explicit knowledge are operationalized and can be measured as outcomes from second language instruction (Ellis, 2005b: 152).

Ellis proposes seven key determinants which intend to serve as basic categories of how implicit and explicit knowledge can be located on a scale: "Degree of awareness, Time available, Focus of attention, Systematicity, Certainty, Metalanguage, Learnability" (2005b: 151-152). To connect these aspects with implicit learning, learners should be given less amount of time in order to "response to something according to their feel[ings]" (ibid.: 152.). Implicit knowledge can be made visible in tasks where the main objective is of communicative value, and where learners do not have to rely on their knowledge of certain terminology in the target structure (ibid.). Regarding the learners' responses, implicit knowledge is when certain structures are accurately produced in a routine manner without struggling, even if the response might be grammatically inaccurate (ibid.).

In contrast to the operationalization of implicit knowledge, explicit knowledge is applied when learners consciously refer to certain "rules without being under time pressure" (ibid.). They have not acquired enough implicit knowledge (in terms of the definition previously given) therefore their responses will remain inconsistent in terms of accuracy while also showing some amount of uncertainty in their responses depending on their "metalinguistic knowledge" (ibid.).

2. The Interface Hypotheses

Having introduced and defined the terms of explicit and implicit knowledge, it is worth mentioning whether or not they are related to each other and, if so, in what kind of manner. These relations can be summarized under the term of "interface" (Hulstijn and De Graaff 1994: 98). Three dominant positions of possible relations can be discovered in the literature on explicit and implicit learning. The "non-interface position" (ibid.) argues that both types of knowledge are isolated systems and therefore explicit knowledge of the L2 cannot be transferred into implicit, procedural knowledge (which the definition previously given implies) (ibid.). Supporters of this interface position argue that there may be neurological reasons involved, that is, that the storage of explicit and implicit knowledge are located in different parts of the brain (Ellis 2005c: 37). Moreover, proponents, who suggest that the second language is acquired in the same manner as first language, argue that children who acquire their native language through predominantly unconscious processes do not have great capacities of explicit knowledge of their first language (Hulstijn and De Graaff 1994: 98).

Another core theory of an interface which potentially exists between explicit and implicit knowledge has been proposed by Ellis (2006: 97). According to his "weak interface position" (ibid.), explicit knowledge can support the development of implicit knowledge. This development occurs if the learner is actively involved in the learning process through being provided with language material designed to draw the learners' attention to specific target language structures (ibid.). However, it is questionable whether or not providing learners with explicit knowledge about a certain structure which is beyond their current "stage of interlanguage development" (De Graaff and Housen 2009: 728) does affect the formation of implicit knowledge and therefore the acquisition of a second language (ibid.). According to Ellis learners can benefit from explicit knowledge if they have reached a certain level of readiness in order to be able acquiring the target structure (2006: 97).

The weak interface hypothesis is connected to the concept of the importance of learning with attention and awareness (Leow, 2015: 70). This is in accordance with the "noticing hypotheses" (ibid, see also Schmidt, 1994: 166 for empirical data) which underpins the importance of consciously learning something in the target language. The learners also have to have an awareness of what they are learning (ibid.). Learning without being aware of what is being learned (or taught) cannot be supported through empirical evidence that its effects are more successful in comparison to explicit learning with some amount of awareness (Leow, 2015: 198). When learners are aware of what they are intended to learn, given adequate learning opportunities, their recognition and production skills are also involved, so there occurs a continuous relation between steps of noticing, cognitive processing and finally language production (ibid.: 198). Awareness can also be raised through using diverse and adapted feedback strategies which can be a basis for "consciousness raising tasks" (Ellis, 2005a: 214). Learners could also use their explicit knowledge for "monitoring their output" (Ellis, 2010: 9) in order to increase their rate of accurate speech while they are speaking (ibid.)

The third interface hypothesis assumes that explicit knowledge becomes implicit if the learners are given enough opportunities to apply newly learned target structures under different conditions and use them repeatedly (Hulstijn and De Graaff, 1994: 99). Such an interface is connected with "skill acquisition theory" (DeKeyser, 2013: 48) and proposes that firstly the learners have some kind of "factual knowledge" (ibid.) of a target structure, which can be verbalized (ibid.). This knowledge is transferred into concrete actions; for example, by answering the question as to what concrete steps need to be undertaken if some kind of problem needs to be solved (ibid.). The knowledge enters the long-term memory and then learners do not have to continually consciously rely on their explicit knowledge because this no longer requires any controlled access (ibid. 48-49).

In addition, potentially losing some type of knowledge which can be verbalized does not mean that learners will also lose their ability to use the target structure accurately (ibid.: 49). "The essential notion to bear in mind here is that proceduralization is achieved by engaging in the target behavior, or procedure, while temporarily leaning on declarative crutches" (ibid, 49). Implicit knowledge can also become explicit when, for example, learners attempt to articulate their knowledge (DeKeyser, 2011: 315).

These three different perspectives of potential interface between explicit and implicit knowledge have important theoretical and methodological implications for theorists and especially for language teachers: All three hypotheses can be supported by empirical evidence, therefore practical use of specific instructional approaches in second language classrooms are affected by the beliefs of teachers or experts in terms of, for example, grammar teaching (Ellis, 2010: 8). Proponents of the non-interface hypothesis will probably choose to give less instruction than teachers who support the interface hypothesis and who will probably use more explicit teaching of grammar rules (ibid.).

Further methodological considerations are presented in Section 6 which discusses the dichotomy between inductive and deductive teaching.

However, at this point, it is important and necessary to debate how instruction can increase the learners' proficiency in a second language by also considering the role of implicit teaching. In general, using "form-focused instruction", according to Ellis (2002: 223), facilitates and enhances learning and acquisition of the foreign language (ibid.).

3. Form-focused instruction vs meaning-focused grammar instruction

De Graaff and Housen (2009: 736) define form-focused instruction as

any instructional activity which aims at drawing learners' attention to language form, where "form" stands for grammatical structures, lexical items, phonological features and even sociolinguistic and pragmatic features of language.

This term can also be divided into the sub-terms: "focus on form" (Ellis, 2002: 225) and "focus on forms" (ibid.). Whereas the latter approach primarily focuses on teaching numerous planned grammatical features in isolation, and excludes meaningful activities, focus on form does contain a "context of communicative activity" (ibid.). Norris and Ortega identify six core characteristics of a focus on form approach (i.e. "(a) designing tasks to promote learner engagement with meaning prior to form, (b) seeking to attain and document task essentialness or naturalness of the L2 forms; (c) attempting to ensure that instruction was unobtrusive; (d) documenting learner mental processes ("noticing"); (e) selecting target form(s) by analysis of learners' needs; or (f) considering interlanguage constraints when choosing the targets of instruction and when interpreting the outcomes of instruction") (2000: 438).

These principles can be summarized through stating that target structures should be embedded in meaningful and contexts which activate the learners without an intensive focus on form and also consider the individual differences among the learners (ibid.: 438). Loewen describes the danger an intensive focus on form entails, namely that the communicative focus of a lesson or a task is inhibited through a shift from the primary focus on meaning to an increased amount of negotiation about accuracy (2015: 58). According to Thornbury,

A focus on form does not necessarily mean a return to drill-and-repeat type methods of teaching. Nor does this mean the use of an off-the-shelf grammar syllabus. A focus on form may simply mean correcting a mistake (2009: 24).

Language teachers have several opportunities to integrate a focus on form into the classroom. A purposeful focus on form can arise during communication where the teacher might decide to shift away from meaning to form or s/he could plan ahead to incorporate the form the learners will attend to in a certain lesson (ibid.: 59). While learners are communicating and practicing the language, the teacher or their interlocutors will perhaps interfere with the learners' oral production and provide some type of feedback in order to focus on accuracy (ibid). "Immediate feedback instead of delayed feedback" (ibid.) is one of the core components of this theory because of the less communicative setting that learners are located if they receive some delayed feedback, because it may not be situated in a communicative context anymore (ibid.).

A well- functioning and learner-centered focus on form combines the formation of both types of knowledge (i.e. explicit and implicit) as well as both types of learning (i.e. explicit and implicit), therefore numerous potential skills are practiced and the learners build more resources they can further use in communication (Loewen, 2015: 57).

The advantage of using a focus on form approach, in the communicative language classroom, compared to either focusing on target forms or on meaning only is that the actual communicative function of a certain structure becomes visible to the learners (Ellis, 2002: 225). They can benefit from this shift especially while they are working on communicative tasks where some "negotiation for meaning" (Long and Robinson, 2013: 22) is involved. If teachers draw the learners' attention to certain forms, they should give them a number of opportunities to communicate with each other (ibid.). The interlocutors serve as a resource for providing the learners with feedback about their speech (ibid.). When the interlocutor is the teacher, or another perhaps more proficient speaker, s/he could use recasting techniques, that is, accurately stating the message the learner intended to deliver (ibid.: 23). Several other opportunities for providing feedback exist and they range from asking the students to notice any gaps between their produced speech and the speech by their interlocutor to more explicit focus on form through correcting errors using metalinguistic language (Ellis 2006: 100).

Ellis identifies some potential problems which can arise when teachers aim to develop implicit knowledge among their learners through teaching that includes a focus on target forms: Learning and acquisition of a second language demand a great amount of time while learners have limited amount of opportunities for practice (2002: 224). Therefore learners' foundation of implicit knowledge from the depth of cognitive processing of a certain structure cannot be guaranteed (ibid).

However, approaches that predominantly focus on meaning are characterized through the principle that "focal attention is predominantly on the communication of relevant meanings and authentic messages" (De Graaff and Housen, 2009: 735). Such teaching principles can be identified as, for example, "communicative language teaching (CLT), immersion programs or content and language integrated learning" (CLIL) (ibid.).

Focus on forms involves a number of potential problems for the language learner: perhaps the presented target structures do not have any relevance, for the learners, as to why they should be learned (Long and Robinson, 2013: 16). Such models or methodologies, for example, the "Grammar Translation Method" (ibid) have not shown evidence that there is an increase of language learning (ibid.). The improvement of the learners' abilities to communicate, to interact and to express certain meaning in the target language, remains questionable when teachers apply such approaches (Motha, 2013: 6). On the other hand, approaches and methods with a dominant focus on meaning which is connected to the above mentioned theories about implicit teaching and implicit acquisition of the second language also consist of gaps in preparing the learners to develop and produce accurate language (ibid.: 7).

Long and Robinson also discuss the learners' differences in age as having an effect on their language acquisition which can result in a potential decrease of the abilities of implicit learning (2013: 19). Especially older learners are endangered of losing these skills over a certain period of time (ibid.). Younger learners show greater capabilities to implicitly learn a second language through being intensively exposed to it, but unless there is a focus on form during their time spent on acquisition, their language production will remain inaccurate (ibid.: 20). On the other hand, although adult learners can also benefit from intensive exposure, in comparison to younger learners they probably will not reach the same levels of proficiency (ibid.).

Another important factor which causes difficulties of acquiring target structures through an isolated focus on meaning occurs when, for example, target structures are not present in the native language of the learner (ibid.). In most cases, the disruption when communicating is limited, but speech has enormous gaps and errors (ibid.: 21).

The amount of influence of cognitive processes necessary for second language acquisition causes the need or the requirement that learners, who are willing to comprehend and further process new target structures, are somehow dependent on the teacher whose primary aim should be to facilitate these developments (ibid.: 20). In contrast to this, more traditional explicit approaches (i.e. "translation" [ibid.]) have not shown that the learners have reached the intended nativelike competencies, therefore meaning focused instruction serves as an alternative in second language classrooms (ibid.).

4. Explicit Instruction

As already stated, explicit instruction involves the explanation of an underlying rule of a certain language pattern or the guiding of the learners to produce their own rule generalizations (Norris and Ortega, 2000: 437). Ellis (2005; cited in De Graaff and Housen, 2009: 726) explains that explicit instruction "may not be necessary to achieve competence in the L2, it undoubtedly helps." Norris and Ortega argue that "L2 instruction can be characterized as effective in its own right" (2000: 480). The following paragraphs attempt to investigate which underlying concepts and theory support these views of explicit instruction.

De Graff and Housen identify three core components of what explicit instruction should contain regarding the perspective of second language acquisition: learners should be trained to "internalize new L2 features, modify L2 knowledge and consolidate L2 knowledge" (2009: 731).

Comparing the effectiveness of either explicit or implicit instruction on learners' outcomes, in their meta-study, Spada and Tomita report an overall advantage for explicit instruction in terms of more complex language features which revealed the largest effect on learners' knowledge (2010: 281). Explicit instruction of more simple language features has also shown greater effects than implicit instruction on complex language features which indicate that regardless of the complexity of the taught structure, explicit instruction results in higher knowledge (acquisition) of the target structures (ibid.). In addition, explicit instruction shows high effect sizes in the learners' abilities to spontaneously respond to requirements of the previously taught complex language structure (ibid.: 287).

On the other hand, implicit instruction also showed effects on the learners' knowledge of the second language and on their production, but not to such an extent as the learners who received an explicit treatment (ibid.: 290). According to this study, the effect of explicit instruction is related to the interface hypothesis because the learners were able to transform their explicitly received knowledge into implicit knowledge by being given numerous opportunities to practice the new target structure (ibid.: 287). It is argued, however, that implicit instruction does not activate the learners' procedural, implicit knowledge of the second language and can therefore not result in greater degrees of free language production (ibid.: 288).

However, the study by Spada and Tomita (2010) needs to be taken with a degree of caution because a few studies in this meta-analysis measured spontaneous speech production by asking the learners to describe some pictures (ibid.: 287) where the learners were not given much time to organize their speech in advance, aiming to diminish the influence of explicit knowledge and eventually leading to doubtful results of measuring implicit knowledge (ibid.). Plus, potential effects of implicit instruction may require a greater amount of time to become visible in the learners' interlanguage, a fact which also potentially influences the results of the analyzed studies (ibid.).

The effectiveness of explicit instruction on spontaneous language usage remains debatable because of the strong evidence of developmental stages among the learners, but the provision of explicit treatment of a target structure facilitates the acquisition once the learners have reached a certain stage of development (De Graaff and Housen, 2009: 728). One important problem explicit instruction can have for the learners is the contradiction between the "assumption that language is comprised of a system of grammatical forms and structures that can be acquired consecutively" (Loewen, 2015: 85). Another problem is related to the procedures undertaken in second language classrooms, where the explicitly instructed language features do not represent current developmental stages of the learners, but lay above them and therefore cannot be acquired (ibid.).

However, De Graaff and Housen argue there is an existing gap between explicit instruction and the ability to produce fluent language, whereas the learners' language accuracy can be improved through explicit treatment (2009: 729). This finding still shows one advantage of explicit instruction over implicit instruction because the more explicitly teachers instruct their learners, the more precisely learners will produce language in terms of lexical correctness and "sociolinguistical appropriateness" (ibid.: 735). Concerning further aspects of the effectiveness and success of explicit instruction of second language features, different empirical results are presented in the literature (ibid.: 740).

Spada and Tomita report an advantage of explicit instruction on more complex features of the language (2010: 281), and De Graaff and Housen argue that more simple structures, in terms of the learners' levels, are positively affected through explicit instruction (2009: 740).

Since both studies report advantages of explicit over implicit instruction concerning different levels of complexity among specific target structures, it could be concluded that explicit instruction is more effective and the actual complexity does not show negative influence on the learners' results of acquisition after they have received explicit treatment (Spada and Tomita, 2010: 287). Explicit instruction of target structures is connected with increasing language proficiency, that is

learners acquire both a rich repertoire of formulaic expressions, which cater to fluency, and a rule-based competence consisting of knowledge of specific grammatical rules, which cater to complexity and accuracy (Ellis, 2005a: 210).

Reaching these objectives through explicit instruction is possible if the teachers' instruction focus on "semantic and pragmatic meaning" (ibid.: 211). The first concept includes the formation of awareness of, for example, the meaning of "lexical items" (ibid.), and pragmatic meaning sensitizes the learners to the "meaning in conversation" (ibid.) of certain language patterns.

Loewen (2015: 86) raises the issue of the learners' preference of being either explicitly or implicitly instructed. Adult learners, in particular, prefer explicit instruction because of their already acquired experience of how learning environments function and they view knowledge about specific rules as a facilitating factor in their acquisition processes (ibid.). In addition, older learners are not as able as younger learners to make use of their implicit learning abilities, so a deeper focus on explicit instruction is required in order to increase their language proficiency (ibid.).

However, Hulstijn proposes taking into account the available pre-knowledge of older learners (1995: 366). If they do not have great proficiency in their second language plus a limited amount of time available to learn, explicit instruction of grammar rules does not lead to increasing second language proficiency (ibid.). The same observation can be made for "illiterate or literate learners who have only been to elementary school" (ibid.). Demanding and asking those learners to learn a language through being explicitly taught certain grammar rules negatively influences the motivation and self-confidence to continue with second language learning (ibid.; see also Ellis, 2006: 91).

Concerning the age of learners and the degree of explicitness in teaching new language structures, Ellis discusses two concepts, that is, grammar being taught explicitly to younger learners or not (2006: 90): He argues that, in terms of the interface debate, by explicitly teaching learners, they can derive their explicit knowledge and, later, transfer this into implicit knowledge and this then facilitates their later usage of certain grammar structures (ibid.).

Ellis also describes the benefits of immersion programs for younger learners, in particular, because the general levels of proficiency appear to somewhat high (ibid.) even though these learners show significant gaps in terms of accurate production when compared to native speakers of that language (Long and Robinson, 2013: 21). Therefore, a later beginning of explicit teaching of grammar, after learners have already acquired certain communicative skills, by including aspects of "TBLT" (Ellis, 2006: 91) for example, are more beneficial for learners' proficiency (ibid.). A delayed beginning of explicitly teaching grammar also considers the language level of beginners who attempt to convey their communicative message without paying conscious attention to their form (ibid.). Building communicative competencies at the beginning of the acquisition processes combined with an overall focus on meaning rather than on accuracy also show greater impacts on long term interlanguage developments (ibid.).

The main questions that remain are: Which language features can be learned without receiving explicit instruction and which features require being explicitly taught? Are there other opportunities to activate learning processes through, for example, different strategies of feedback (Ellis, 2006: 90; 103)? Explicit instruction does not reject the idea of providing the learners with implicit feedback strategies and especially "recasts" (Ellis, 2010: 7) can be beneficial for learners because they do not need to change their message conveyed (ibid.). "A corrective recast reformulates the learner's erroneous utterance with the correct form highlighted intonationally" (ibid). Feedback strategies are thus connected with the complexity of the chosen target structure, the more demanding the structure and therefore eventually taught more "intensively" (Ellis, 2006: 94), the less chance there is for the learners to notice their interlanguage gaps (ibid.). The teacher will address their feedback more on the accurate usage of a structure if the lesson's primary objective is the accurate production of this previously taught structure.

Doughty and Williams raise the issue of whether adult learners are able to access their "Universal Grammar (UG)" (2013: 201) that is, their "system of categories, mechanisms and constraints shared by all human languages and considered to be innate" (Dabrowska, 2015: 2). If adult learners lose the capacities to access their UG, the more explicit instruction will be necessary in order to reach higher levels of proficiency among these learners (Doughty and Williams, 2013: 201). If this is not the case, such structures which are still part of the UG do not need to be taught in an explicit manner, the occurrence in the language input would be sufficient enough to ensure that learning these structures could be initiated and established (ibid.).

5. The Target Structure

According to Ellis an existing problem of measuring the actual outcomes of second language instruction and its impact on communicative usage of new grammatical forms prevents research from providing clear assumptions about which language structures should be taught (2006: 87). Designing an appropriate teaching sequence in terms of grammar teaching requires the selection of a grammar structure which is useful to teach and distinguishable from those which do not have to be taught (ibid.: 88). The teacher has to decide whether or not the chosen structure is easy or difficult to comprehend, for example if the structure itself causes potential difficulties in terms of being comprehensible for the current level of the learners (ibid.). The second issue for the learners describes the "difficulty they have in internalizing a grammatical feature so that they are able to use it accurately in communication" (ibid.). On the other hand, when such difficulties in comprehending a new structure are not expected from the learners, "why bother to teach what can be learned naturally?" (ibid.: 91). According to Robinson, a rule contains "a claim about the form in which knowledge of language is represented in the learner's mind" (1996: 31).

In addition, "pedagogic rules" (ibid.: 32) are intended to present the learners a new target structure in a more comprehensible and simplified manner and do not contain complex linguistic explanations (ibid.). Learners can benefit from these because their attention and the main components of such rules refer to more conspicuous language features (ibid.). These are then, ideally, embedded in meaningful examples which represent the form and its meaning (ibid.). The teacher also has to ensure that his/her explanation does not contain language that is too complex and therefore incomprehensible for the learners (ibid.).

DeKeyser argues that, for example, new vocabulary can be acquired through learners being exposed to the target language, whereas specific grammar is more difficult to grasp from the input only (2013: 43). The problematic issue of a structure being eventually accessible through Universal Grammar raises the question whether a more intense focus on those forms is necessary (ibid.). Besides, "negative evidence" (ibid.) of a structure may help its acquisition, especially if the teacher aims to distinguish the structure between the learners LI and L2 (ibid.). Answering this question also requires more knowledge about which specific grammar aspects are easy to access for the learners, in terms of their Universal Grammar and which ones do cannot be located within this system (ibid.).

Another main problem of identifying what language structures should be taught is the lack of agreement of how to define the term "complexity" (ibid.: 44). DeKeyser refers to Krashen's definition of distinguishing "between rules that are easy to acquire but hard to learn, and rules that are easy to learn but hard to acquire" (2013: 43) meaning that a greater degree of explicit teaching of rules which tend to be more difficult to be internalized is necessary for the learners' improvements in the target language (ibid.). Depending on the focus of what kind of theoretical approach is considered, one could argue that learning the underlying rule of the usage of "third person -s" (ibid.: 44) is, in Krashen's terms rather simple because of the less necessary explicit instruction in order to increase the language proficiency (ibid.). When considering the theory of developmental stages, the exact opposite view is favored (ibid). "The point is that, instead of giving up on more difficult rules, teachers may have to put the most emphasis on them" (ibid.).

Hulstijn and De Graaff also argue in favor of explicit instruction when the focus of the lesson is placed on more complex rules which are characterized through greater cognitive demands on the learners (1994: 103). They refer to such rules of language structures which are not obvious or "salient in the language input" (ibid.). Importantly, explicitly teaching such rules also means having more time to practice them (ibid.).

Defining and operationalizing the complexity of a certain language rule also requires the consideration of further terms which provide the framework of a deeper understanding of complexity (Spada and Tomita 2010: 266). The term "psycholinguistic complexity" (ibid.) refers to Pienemann's developmental stages of second language acquisition, stating that a certain stage can only be mastered if the previous stage has also been mastered (ibid.). The problem connected with the selection of rules to teach is that, according to this theory, the more complex the rules are, the later the learners are able to acquire them (ibid.).

[...]