The Best of Both Worlds: Combining Cognitive Linguistics and Pedagogic Tasks to Teach English Conditionals

The Best of Both Worlds: Combining Cognitive Linguistics and Pedagogic Tasks to Teach English... Abstract Due to their internal complexity, English conditionals represent a challenge for L2 learners; furthermore, most ESL grammar books lack precise explanations of conditional meanings and usage contexts. As a solution, cognitive linguistic research on conditionals ( Dancygier and Sweetser 2005 ), based on theory of conceptual blending and mirroring human cognition processes, could be very beneficial for L2 learners. In this study, elements of cognitive linguistic analysis of English hypothetical conditionals were incorporated into L2 learning materials. The design of the study was quasi-experimental, with three groups ( N = 57) participating: cognitive, task-supported, and control. Pedagogic consciousness-raising tasks were used to provide a context for meaningful practice for cognitive and task-supported groups. The research questions explored the efficacy of different types of instruction for L2 development of English conditional phrases, as measured by a post-test and a delayed post-test. The findings of t -test and analysis of variance procedures ( p = .001 ) indicate that participants from the cognitive group outperformed the participants from the task-supported and control groups and demonstrated a higher degree of understanding of conditional usage contexts. INTRODUCTION The value of using cognitive linguistics for language teaching Cognitive linguistic (CL) theory of language and usage-based approaches to language pedagogy have been gaining recognition by second language acquisition (SLA) researchers ( Robinson and Ellis 2008 ; Tyler 2012 ) over the course of the past couple decades. CLs assumes that language is composed solely of form-meaning pairings, or ‘symbolic units’ ( Langacker 2008 ). Under this analysis, syntax and morphology cannot be strictly separated from lexis, but rather, all linguistic units carry some sort of meaning. Investigating the meaning of symbolic units can be equated with exploring the nature of conceptualization. Ways to conceptualize a specific event or entity, and thus the message, are inherently subjective, as they vary widely depending on the speaker (his/her background knowledge and choice of discourse focus) and the assessment of what the listener already knows. Accordingly, to explore the meaning of a language expression, one would need to determine the conventional linguistic choices, the context of use, and the background knowledge it appears to evoke. CL view of language thus allows to uncover the semantic motivation behind certain syntactic, grammatical, or morphological phenomena. Using select, carefully adapted for unprepared audiences, and contextualized CL concepts in L2 instruction can reveal the perspective of a native speaker and make form-meaning mappings relatively transparent for L2 learners. Accordingly, the underlying conceptual characteristics of CLs make it a good candidate for the role of providing a comprehensive theory that could successfully support L2 instruction ( Achard and Niemeier 2004 ; Tyler and Evans 2004 ; Tyler 2012 ). A number of researchers conducted studies using CL theory to inform English L2 teaching. CL approaches were successfully applied to teaching of phrasal verbs ( Dirven 2001 ; Liu 2010 ), metaphorical/idiomatic language ( Lindstromberg and Boers 2005 ; Littlemore and Low 2006 ; Boers & Lindstromberg. 2008 ; Yasuda 2010 ), and grammar (prepositions: Lindstromberg 1996 ; Tyler and Evans 2004 ; Tyler Mueller and Ho 2011 ; tense and aspect: Niemeier and Reif 2008 ). (For a more detailed overview of these and other studies, see Jacobsen 2012 , 2015 .) Despite the rich qualitative support for adapting CLs to L2 instruction, the existing number of experimental and/or quantitative research studies is not sufficient to provide solid empirical backing to L2 instructional benefits associated with CLs: that is, some of the above-mentioned studies did not employ a rigorous research design or use statistical procedures that would allow for corresponding findings to be considered generalizable. Also, there is a dearth of research into how CL L2 applications can be paired or combined with best practices in language pedagogy. For instance, Achard (2004) examined CL pairing with the Natural Approach, while Holme (2009) explored CL applications of total physical response (TPR) to the L2 classroom. While both of these pedagogical approaches are valid for a limited range of language structures, it is challenging to apply them in all language-teaching contexts. Boers (2013) reached a similar conclusion about the limits of L2 applicability of CL metaphor theory: while concepts from metaphor research have been helpful for teaching figurative language, existing applied CL research does not promote learner recognition and retention of new figurative phrases, nor does it overcome the effects of learners’ individual differences, such as motivation, proficiency, and learning styles. In sum, the generalizability of CL advantages for L2 instruction still appears to be limited, and much more research needs to be conducted to further validate or possibly refine the previously established benefits. The goal of the present study is to investigate the relevance of CL for L2 teaching, addressing the above-mentioned gaps as much as possible. The study focuses on the efficacy of applying CL analysis of English conditionals to L2 instruction, using a mixed-methods design and supplementing CL approach with task-supported language teaching. The concept of conditionals: Why are they problematic for L2 learners? Conditional constructions allow humans to negotiate between several logical scenarios and thus capture various consequences of their actions or relevant circumstances, including making predictions about events that have not yet happened in the present or that did not happen in the past. The surface form of English conditional constructions is frequently referred to as if P, then Q , where P is the conditional clause, and Q is the head clause ( Comrie 1986 ; Jackson 1991 ; Woods 1997 ; Evans and Over 2004 ). The P-clause sets up a ‘theoretical’ situation and marks it as remote from actuality, the Q-clause takes it on to a further outcome, marking it with some degree of probability ( Werth 1997 : 252). The classification most widely used in ESL contexts is provided in Table 1 below. Table 1: Classification of conditional forms ( Taylor 1997 ) Type Example Meaning and typical form Factual If prices go up, I sell my car. Content of the if -clause is presumed to be real and true. Form : present tense in the if -clause and present/future tense in the main clause Hypothetical If prices went up, I would sell my car . Content of the if -clause is regarded as a possibility. Form : past tense in the if -clause and a modal such as would (or similar form) plus a base verb form in the main clause. Counterfactual If prices had gone up, I would have sold my car . Content of the if -clause is regarded as not possible or contrary to the fact/current state of the world. Form : past perfect in the if -clause and a would (or another modal) combined with the perfect form of the verb in the main clause. Type Example Meaning and typical form Factual If prices go up, I sell my car. Content of the if -clause is presumed to be real and true. Form : present tense in the if -clause and present/future tense in the main clause Hypothetical If prices went up, I would sell my car . Content of the if -clause is regarded as a possibility. Form : past tense in the if -clause and a modal such as would (or similar form) plus a base verb form in the main clause. Counterfactual If prices had gone up, I would have sold my car . Content of the if -clause is regarded as not possible or contrary to the fact/current state of the world. Form : past perfect in the if -clause and a would (or another modal) combined with the perfect form of the verb in the main clause. Note : Examples come from Werth (1997 : 243–5). View Large Table 1: Classification of conditional forms ( Taylor 1997 ) Type Example Meaning and typical form Factual If prices go up, I sell my car. Content of the if -clause is presumed to be real and true. Form : present tense in the if -clause and present/future tense in the main clause Hypothetical If prices went up, I would sell my car . Content of the if -clause is regarded as a possibility. Form : past tense in the if -clause and a modal such as would (or similar form) plus a base verb form in the main clause. Counterfactual If prices had gone up, I would have sold my car . Content of the if -clause is regarded as not possible or contrary to the fact/current state of the world. Form : past perfect in the if -clause and a would (or another modal) combined with the perfect form of the verb in the main clause. Type Example Meaning and typical form Factual If prices go up, I sell my car. Content of the if -clause is presumed to be real and true. Form : present tense in the if -clause and present/future tense in the main clause Hypothetical If prices went up, I would sell my car . Content of the if -clause is regarded as a possibility. Form : past tense in the if -clause and a modal such as would (or similar form) plus a base verb form in the main clause. Counterfactual If prices had gone up, I would have sold my car . Content of the if -clause is regarded as not possible or contrary to the fact/current state of the world. Form : past perfect in the if -clause and a would (or another modal) combined with the perfect form of the verb in the main clause. Note : Examples come from Werth (1997 : 243–5). View Large In traditional instruction, learners are presented with three conditional types—factual, hypothetical, and counterfactual—and corresponding examples thereof. A number of factors contribute to the difficulty of conditionals for L2 learners ( Celce-Murcia and Larsen-Freeman 1999 ). First, because the conditional structure consists of two clauses, the syntactic aspects of subordination do not make them easy to acquire. Second, before being able to form conditionals correctly, learners need to master a number of other grammatical features, such as tense and aspect, modal auxiliaries, and negation. Lastly, the variety of types and contexts in which conditionals are used are often inadequately captured by typical ESL/EFL sources, as described in the following section. Representation of conditionals in EFL/ESL materials: Common problems In the majority of current ESL/EFL textbooks, the aspects of conditional form—as opposed to meaning—represent the prevalent instructional focus ( Jacobsen 2012 ). A typical classification (as referred to in Table 1 above) and accompanying examples do not explain how conditionals may be used in a variety of contexts, nor do they provide a clear link between different conditional types and their corresponding semantic motivations, or the exact linguistic environments where one type of conditional might be preferred over others ( Meunier 2008 ). (For a full overview of how conditionals are presented in textbooks, see Jacobsen 2012 ). Furthermore, a recent study based on corpus findings (Gabrielatos 2003, cited in Meunier 2008 ), that compared the traditional English language teaching terminology of conditional clauses against corpus evidence, found that 55 per cent of the authentic corpus examples did not fit into the core english language teaching (ELT) typology. Examples produced in naturally occurring contexts were found to be much more varied and multi-faceted than those captured by the traditional typology. For instance, using would forms in if -clauses is often treated as incorrect in textbooks; however, corpus data from different registers show that such usage is indeed permissible. In other words, traditional typologies appear to be at best schematic, and misleading and imprecise at worst, but still continue to be widely used in ESL/EFL materials. To counter these shortcomings, the CL approach has the tools to address conditional meaning and context usage, and make them fully observable and salient for the learner. CL approach to conditionals This study uses the CL analysis of English conditionals by Dancygier and Sweetser (2005) . Basing their approach on blending theory by Fauconnier and Turner (2002) , Dancygier and Sweetser claim that conditional if -clauses set up certain mental space structures. They analyze how such a set-up happens through manipulating various markers (e.g. if , when , unless ) and using different verb forms. Coherence and validity of conditional phrases depend on the successful configuration of all the elements constituting their structure, as individually construed by a given speaker. The authors’ approach takes into account the contextual needs and formal characteristics of conditionals as they fit in each context of use. Thus, Dancygier and Sweetser outline linguistic tools commonly employed by native English speakers to create felicitous conditional phrases, while also demonstrating the dependence of conditionals on the communicative context and the speaker’s stance toward a given situation. An important insight from Dancygier and Sweetser’s work is the idea of compositionality of conditionals: by analyzing combinations of regular and less regular conditional constructions and identifying patterns of inferential structure and metonymic reasoning involved, the authors essentially single out the elements constituting conditional constructions and make those aspects of meaning directly analyzable. Treating conditionals as meaning-form ‘packages’ ( Achard 2004 ) allows us to break down composite grammatical meanings into separate parts and subsequently highlight the form–meaning connections of those parts explicitly to learners. A careful adaptation of these CL insights into L2 materials has the ability to highlight individual aspects of conditional meaning and also outline patterns that learners could readily rely upon in the subsequent construction of their own conditional sentences. Explicit instruction and task-based language teaching CL theory can be incorporated into the classroom through explicit instruction, the effectiveness of which has been firmly established in prior research ( Norris and Ortega 2000 ; Ellis 2002 ; Spada and Tomita 2010 ). In particular, the underlying tenets of focus on form ( Long 2000 ), noticing ( Schmidt 2001 ), and consciousness-raising ( Sharwood Smith 1981 , 1993 , 2007 ) emphasize the effectiveness of helping the learner notice and understand key meaning motivations behind target forms. Focus on form or consciousness-raising is instrumental in helping learners notice linguistic cues, especially when targeting complex, polysemous, or potentially ambiguous linguistic forms (cf., Hulstijn & de Graaff 1994 ; Fordyce 2014 ; DeKeyser and Prieto Botana 2015 ). Furthermore, proponents of CL theory have suggested combining CL principles with selective focus on form ( Achard and Niemeier 2004 ; Cadierno 2008 ; Holme 2012 ). However, viable pedagogical solutions are needed to make these novel insights appear approachable to the learners. The paradigm of task-based (-supported 1 ) language teaching (TBLT) is highly compatible with the basic CL assumptions, the two providing a rich initial ground for further collaboration. TBLT has been gaining increased popularity over the course of the past two decades (e.g. Bygate et al. 2001 ; Robinson 2001 , 2005 ; Ellis 2003 ; Samuda and Bygate 2008 ), having emerged as a natural progression from communicative language teaching and shifting the pedagogic focus onto meaning-oriented and learner-centered instruction ( Van den Branden et al. 2009 ). While there exist a number of definitions for tasks, this study will rely on the comprehensive operationalization of task, provided in Ellis (2003 : 5–6): ‘A task is a workplan that requires learners to process language pragmatically in order to achieve an outcome… it requires them to give primary attention to meaning and to make use of their own linguistic resources, although the design of the task may predispose them to choose particular forms.’ Tasks provide learners with an opportunity to produce language in a context that resembles or aims to recreate an authentic language acquisition environment ( Révész 2009 : 438). While the focus on communication and meaning remains the primary element of TBLT, most researchers agree that focus on form incorporated into instruction in line with the learners’ developmental needs is crucial for boosting L2 acquisition ( Ellis 2003 ; Long 2000 ; Norris and Ortega 2000 ; Robinson 2001 ; Van den Branden et al. 2009 ). The value of using task-based methodology in combination with CL insights has been explicitly addressed in Cadierno (2008) . Furthermore, Cadierno and Robinson (2009) used pedagogic tasks to measure how the manipulation of cognitive complexity can facilitate the development of L2 construal patterns. Moder (2010) used targeted pedagogic tasks (based on prior corpus-based research) to teach usage patterns of the English like constructions. Finally, a number of experiments conducted at Georgetown University reported in Tyler (2012) used various versions of pedagogic tasks to deliver CL insights to learners. While it is a promising direction, the relative dearth of studies at this point in time does not allow us to make comprehensive conclusions about the best ways to combine TBLT with CL insights. This study thus aims to address this gap by complementing the applied CL focus with the use of pedagogic tasks. Task-supported instruction and consciousness-raising tasks The instructional context of this study is closer to ‘task-supported’ rather than ‘task-based’. In task-supported settings ( Ellis 2003 ), tasks are not the main medium of instruction but rather one of the key components of the general syllabus, providing learners with semi-authentic contexts for practicing target forms ( Samuda and Bygate 2008 : 60). As it will become clearer from the ‘Methods’ section, constraints of the graduate English for Academic Purposes (EAP) instructional context made it impossible to re-conceptualize the existing curriculum as entirely task-based. Relying on tasks as means of support, rather than the main component, of instruction was more feasible in the provided educational environment. In the present study, the task support was realized in form of the consciousness-raising tasks . Eckerth (2008a : 92) defines consciousness-raising tasks as ‘form-focused tasks’ that can be used as ‘a pedagogical device to direct learners’ attention to specific L2 forms while they are communicating in the L2’. The main purpose of consciousness-raising tasks is to give learners the opportunity to explore target features in a focused context while simultaneously attending to completing task guidelines. The efficacy of consciousness-raising tasks was proven in multiple L2 contexts ( Fotos and Ellis 1991 ; Fotos 1993 , 1994 ; De la Fuente 2006 ; Eckerth 2008a , 2008b ). Specifically, Fotos (1994) examined how grammar-focused consciousness-raising tasks affected proficiency gains in acquiring word order, finding that consciousness-raising tasks were a more powerful teaching method than formal instruction. In De la Fuente’s (2006) study on the efficacy of using consciousness-raising tasks for instruction of vocabulary, participants of the task-based group performed statistically significantly better on the delayed L2 vocabulary retrieval test than the traditional group participants, which suggests a positive long-term effect associated with consciousness-raising tasks. Eckerth (2008a ) reported that learners’ language scaffolding skills are necessary for successful completion of consciousness-raising tasks. Eckerth’s (2008b ) study focused on investigating the effects of dialogic tasks onto specific learning gains and on the process of task completion itself. The findings demonstrated that task completion yielded significant task-specific learning gains both immediately and after passage of some time. Furthermore, consciousness-raising tasks were associated with greater linguistic complexity in terms of both perception and production. These findings provide support for using consciousness-raising tasks as a way to deliver guided focus on form in L2 settings, that is, helping highlight CL insights in a meaningful usage context. METHODS Target structure The author conducted preliminary research in free online learner corpora (for full list with web addresses, see Appendix A in Jacobsen 2012 ) on most commonly occurring errors in conditionals, finding that the range of constructions carrying the predictive 2 function (e.g. If prices go up, I’ll sell/I sell my car/If prices went up, would sell my car/If prices had gone up, I would have sold my car )—as opposed to the pragmatic/speech act conditionals—represents the biggest challenge for L2 learners. In other words, learners experience the most difficulty (as exhibited through imprecise or erroneous tense sequences and/or other time indicators) when attempting to highlight hypothetical outcomes referring to the unreal situation in the past or the present. Accordingly, predictive conditionals were chosen as the target structure of this study. Research questions and hypotheses This study intended to examine the general effects of task-supported focus on form upon the instruction of conditionals, as well as the relative significance of CL insights included into the instructional treatment. The study addressed three research questions: Research Question 1 : Does task-supported instruction produce an effect for L2 development of conditional constructions? Research Question 2 : Does task-supported instruction with added CL insights produce an effect for L2 development of conditional constructions? Research Question 3 : Does the addition of CL insights into task-supported instruction produce greater L2 development of conditional constructions than task-supported instruction alone? The term ‘ CL insights ’ refers to CL explanations added to pedagogic materials. CL insights were part of instruction for only one of the groups (referred to as the cognitive group hereafter). The group that received instruction without the CL insights will be referred to as the task-supported group. Finally, effect for L2 development is operationalized as performance on the post-test and the delayed post-test (see section on study design). While there does exist considerable research pertaining to the effectiveness of CL in L2 teaching, its total scope and limitations do not yet allow for directional hypotheses; accordingly, it would be most appropriate to frame the research hypotheses as null statements. H1 : Task-supported instruction will not produce any effect for the L2 development of conditional constructions. H2 : Task-supported instruction with added CL insights will not produce an effect for L2 development of conditional constructions. H3 : The addition of CL insights into task-supported instruction will not produce greater L2 development of conditional constructions than task-supported instruction alone. Participants The data were collected in a graduate EAP program at a large university on the East Coast of the USA. The study included three main groups— cognitive , task-supported , and control . The subjects in the cognitive and task-supported groups were recruited from single intact graduate EAP classes, while the subjects in the control group came from two separate EAP classes. Most of the subjects were in their mid-20s. The total number of male participants across three groups was 36, and the total number of females was 21. The vast majority of participants came from Mainland China or Taiwan and spoke Mandarin Chinese as their first language, but there were a few participants with other L1s (see Table 2 ). A key limitation of this population sample is that the influence of L1 on the subjects’ prior conceptualization of the concept of conditionals was not assessed; such an intervention would require logistical arrangements incompatible with the curricular restrictions within the program. However, qualitative analysis of errors in learner corpora demonstrated that learners of all L1 backgrounds still struggle with forming grammatical tense sequences and assuming context-appropriate stance for expressing conditional meaning in English. To ensure that there were no differences in the prior level of mastery of English conditionals, a one-way analysis of variance (ANOVA) was administered on pretest data (see the ‘Results’ section for full report). Table 2: Study participants Cognitive group Task-supported group Control group Total number 17 18 22 Mean age 25 (SD = 5.6) 24 (SD = 4.7) 26 (SD = 4.2) Gender     Male 10 10 16     Female 7 8 6 L1     Mandarin Chinese 14 17 15     Arabic 1 1 4     Farsi 2 1     Portuguese (Brazil) 1     Urdu 1 Mean length of stay in the USA 3 months (SD = 2.3) 3 months (SD = 3.5) 7 months (SD = 2.9) Cognitive group Task-supported group Control group Total number 17 18 22 Mean age 25 (SD = 5.6) 24 (SD = 4.7) 26 (SD = 4.2) Gender     Male 10 10 16     Female 7 8 6 L1     Mandarin Chinese 14 17 15     Arabic 1 1 4     Farsi 2 1     Portuguese (Brazil) 1     Urdu 1 Mean length of stay in the USA 3 months (SD = 2.3) 3 months (SD = 3.5) 7 months (SD = 2.9) View Large Table 2: Study participants Cognitive group Task-supported group Control group Total number 17 18 22 Mean age 25 (SD = 5.6) 24 (SD = 4.7) 26 (SD = 4.2) Gender     Male 10 10 16     Female 7 8 6 L1     Mandarin Chinese 14 17 15     Arabic 1 1 4     Farsi 2 1     Portuguese (Brazil) 1     Urdu 1 Mean length of stay in the USA 3 months (SD = 2.3) 3 months (SD = 3.5) 7 months (SD = 2.9) Cognitive group Task-supported group Control group Total number 17 18 22 Mean age 25 (SD = 5.6) 24 (SD = 4.7) 26 (SD = 4.2) Gender     Male 10 10 16     Female 7 8 6 L1     Mandarin Chinese 14 17 15     Arabic 1 1 4     Farsi 2 1     Portuguese (Brazil) 1     Urdu 1 Mean length of stay in the USA 3 months (SD = 2.3) 3 months (SD = 3.5) 7 months (SD = 2.9) View Large The vast majority of the participants first came to the USA a few months prior to the beginning of data collection; for all students, enrolling in this graduate program was their first educational experience in the English-speaking world. All subjects previously studied English with various degrees of instructional intensity in secondary school and possibly college; however, due to lack of uniform reporting on the subjects’ prior educational contexts, their proficiency was captured only through standardized test scores: everyone earned a minimum of 80 on the TOEFL iBT prior to being admitted to university. Table 2 represents basic biographic information related to the study participants. The treatment types received by each group were as follows: the control group ( N = 22) received no explicit classroom instruction targeting conditionals and only completed the three tests; the task-supported group ( N = 18) received instruction targeting conditionals using the traditional teacher-facilitated explanation and pedagogic tasks, that is, materials for task-supported group did not include explanations guided by CL insights; the cognitive group ( N = 17) received teacher-facilitated presentations of conditionals as well as pedagogic tasks. Teacher-used instructional materials did include CL explanations of the target form. Instructional procedure The entire process of data collection took six weeks. During the first week, all three groups completed pretests. The pretest scores were approximately the same across all three groups (25, 24.5, and 23.5 for the cognitive, task-supported, and control groups, respectively; also see the section on test administration and scoring for details on one-way ANOVA). The treatment for the cognitive and task-supported groups took place over the course of three weeks. Following its completion, immediate and delayed (10 days after the end of instruction) post-tests were administered to measure relative gains that may have happened in the course of the treatment. Each of the tests took approximately 50 minutes to complete. To collect qualitative data in the first week following the completion of treatment, three and four subjects from task-supported and cognitive groups, respectively, were interviewed using the method of retrospective interview . The details and findings pertaining to the retrospective interviews were addressed in full in Jacobsen (2015) and will therefore be omitted here. The full study design is available in Appendix A. Tests The tests created for this study attempted to address both implicit and explicit knowledge and reflected production and comprehension aspects of conditional usage. They consisted of four distinct parts: controlled production, free production, ‘grammaticality judgment/comprehension: pictures’, and ‘grammaticality judgment/comprehension: sentences’. Controlled and free production parts, where students were asked to fill in the blanks and describe a picture respectively, thus demonstrating how they used the target form in specific contexts, aimed to target primarily implicit knowledge ( Norris and Ortega 2000 ; Ellis 2009 ). Grammaticality judgment/comprehension parts, on the other hand, were supposed to target explicit as well as implicit knowledge (cf., Fotos 1994 ; Han and Ellis 1998 ; Ellis 2004 ). Specifically, Ellis (2004 : 249) referred to grammaticality judgment tests as a ‘favored method of investigating L2 explicit knowledge’. The maximum number of points on each test was 49, as demonstrated in the cross-sectional breakdown below: controlled production (13 items, 29 points); free production (4 items, 8 points); grammaticality judgment and comprehension: pictures (4 items, 4 points); grammaticality judgment and comprehension: sentences (8 items, 8 points). Test forms included discourse excerpts representing different genres and registers from the Corpus of American Contemporary English ( http://corpus.byu.edu/coca /) to recreate authentic usage patterns of conditionals. Test images were obtained through searches on Web sites with Common Creative Licenses. Comprehensive answer keys were created for each version of pre-, post-, and delayed post-test. The answer keys reflected the opinions of native speakers obtained during the piloting process. Since most test items were taken from authentic discourse sources, the opinions of native speakers were compared with the original verb tenses from the authentic corpus excerpts; if the sources disagreed, the item was discarded. All test items were then piloted with native and non-native speakers. Only the non-ambiguous items (those not causing difficulty for either native or non-native speakers) were used for the final test forms. Each of the four test subsections had three possible versions (A, B, and C), and those were counterbalanced for each test installment (pretest, post-test, and delayed post-test). Instructional materials Each treatment group received three teacher-facilitated PowerPoint presentations (one per each week of instruction) and six pedagogic tasks (two per each week of instruction). The length of each of the presentations described below was about 50 minutes, or the total of 150 minutes divided across six sessions; in other words, both groups received explicit pedagogical treatment of the same length. However, even though both groups addressed metalinguistic aspects of language, they did so in different ways: that is, each group’s presentation content was condition specific (see next two sections). The pedagogic tasks (see separate section) were the same for both treatment groups. Cognitive group All three cognitive presentations were informed by a single CL account of conditionals ( Dancygier and Sweetser 2005 ) and aimed to represent the CL insights in a learner-accessible manner. The first presentation focused on exposing students to the meaning-centered and compositional reality of language. The corresponding mental steps needed for creating conditionals were presented to students as follows: Start with thinking about the background knowledge in question. What is realistic? What is the possible or projected time of condition ( if -clause)? What is the possible or projected time of result (main clause)? Pick the tenses that will reflect such an arrangement. Check if your sentence makes sense in the given context. (The full version of the first PowerPoint (PPT) is available as appendix in Jacobsen 2015 .) The cognitive chart (available in Appendix B) was designed to serve as a one-stop reminder of all these steps, mirroring the cognitive processes necessary for creating felicitous conditionals. Cognitive group participants could use it during the teacher-facilitated instruction and during the pedagogic tasks that they completed in class. Even though students were encouraged to bring their copies of the cognitive chart to every class, it was noted that not all participants made active use of it. While this supplementary resource was available to everyone, no conclusions can be drawn about its distinct use and/or effectiveness in the context of this study. Furthermore, the second presentation, available in Appendix C, demonstrated the underlying meanings of English tenses and how they are used to express a variety of semantic configurations. The third presentation aimed at showing the subjects how conditionals are shaped within specific usage contexts and how surrounding context-specific information might affect the tense composition of either of the two clauses. The cognitive PPT presentations thus explained the general system behind the notion of conditional sentences by focusing on meaning behind every single grammatical form and its usage contexts. In short, the students had to consider and keep in mind the following characteristics when pondering the compositional meaning and implications behind a given conditional sentence: Tenses in surrounding sentences. Perspective of speaker who is narrating: is it the same person as a given character? Is s/he still within the zone of control or already outside it? Time markers such as adverbs, adjectives, providing an idea of what time is referenced. Appropriateness of forms for corresponding discourse context. Task-supported group The task-supported group received a traditional presentation of the English conditionals informed by L2 textbooks and ‘ESL Grammar Book’ ( Celce-Murcia and Larsen-Freeman 1999 ). During the first week of instructional treatment, learners were given an overview of all three conditional types with subtypes as informed by Celce-Murcia and Larsen-Freeman’s extended classification. During the second week, the focus was on tense combinations in all types of conditional sentences. The tense sequences were highlighted according to the explanations and descriptions provided in the traditional ESL materials. During the third week, the focus was on practicing using these forms in various contexts, as provided by the textbooks. While they received a traditionally balanced presentation of various aspects of the conditional form, subjects of the task-supported group were not asked to engage in conscious reflections on meaning behind different tense pairings beyond the range discussed in textbooks and beyond the meaning context provided by pedagogic tasks. Pedagogic tasks Both treatment groups completed a shared set of pedagogic tasks to support the teacher-facilitated PPT presentations. The conditions of these tasks aimed to resemble the authentic context of target language use as much as possible. All tasks were consciousness-raising in the sense that they directed subjects’ attention toward the use of conditionals, and conditionals were crucial for their successful completion. All tasks and their key characteristics are captured in Table 3 . The results of this instructional intervention are reported in the next section. RESULTS Descriptive statistics One-way ANOVA administered on pretest data ensured that all groups were at the same proficiency level (cognitive vs. task-supported: p = .904, standard error (SE) = 1.6; task-supported vs. control: p = .86, SE = 1.5; cognitive vs. control: p = .60, SE= 1.55). Descriptive statistics for all three groups by testing task and an overview of mean overall scores obtained across three tests by all groups are presented in Tables 4 and 5 , respectively, below. Table 3: Pedagogic tasks used in the study Time Task order and name Knowledge/skills addressed through task Foci Gist of task WEEK 1 1 Sentence strips Sequence of tenses between clauses, time markers Controlled production, comprehension; grammar task In groups of 2–3, students discuss meaning of sentence strips representing two parts of conditionals (with blanks for verb tenses). Then students match if -clauses with corresponding main clauses, supplying the correct verb tenses in the blanks and providing an appropriate rationale for their choice. 2 Understanding background knowledge Speaker knowledge Consciousness-raising, comprehension; grammar task Students are given four conditional sentences and asked to explain background knowledge that had to precede formation of each conditional, tying the rationale to form (tense choices). WEEK 2 3 Hedging in data commentaries Tenses and context knowledge Controlled production, comprehension; grammar task Adapted from Swales and Feak (2004) , this task elicited students’ written data commentary of available statistical information within provided table. Students had to frame their data commentary as hypotheses and predictions using conditionals. 4 Global warming causes Tenses; speaker and context knowledge Comprehension, controlled production; grammar task Students discuss causes and effects of global warming using conditionals. For content knowledge, students relied on materials and class discussions from earlier in the semester. WEEK 3 5 Text repair: break-up letter Tenses, context knowledge, time markers Consciousness-raising, controlled production; communicative task Students identify ill-formed conditionals in a break-up letter and rewrite the letter in own way supplying correct forms based on understanding/interpretation of local semantic context. 6 Seating chart task Tenses, speaker knowledge Free production; grammar task Working in groups, students try to seat people of different and sometimes conflicting backgrounds around one dining table using conditionals. The conditions listed could not be satisfied in their entirety, so students had to try to satisfy as many conditions as they possibly could. Time Task order and name Knowledge/skills addressed through task Foci Gist of task WEEK 1 1 Sentence strips Sequence of tenses between clauses, time markers Controlled production, comprehension; grammar task In groups of 2–3, students discuss meaning of sentence strips representing two parts of conditionals (with blanks for verb tenses). Then students match if -clauses with corresponding main clauses, supplying the correct verb tenses in the blanks and providing an appropriate rationale for their choice. 2 Understanding background knowledge Speaker knowledge Consciousness-raising, comprehension; grammar task Students are given four conditional sentences and asked to explain background knowledge that had to precede formation of each conditional, tying the rationale to form (tense choices). WEEK 2 3 Hedging in data commentaries Tenses and context knowledge Controlled production, comprehension; grammar task Adapted from Swales and Feak (2004) , this task elicited students’ written data commentary of available statistical information within provided table. Students had to frame their data commentary as hypotheses and predictions using conditionals. 4 Global warming causes Tenses; speaker and context knowledge Comprehension, controlled production; grammar task Students discuss causes and effects of global warming using conditionals. For content knowledge, students relied on materials and class discussions from earlier in the semester. WEEK 3 5 Text repair: break-up letter Tenses, context knowledge, time markers Consciousness-raising, controlled production; communicative task Students identify ill-formed conditionals in a break-up letter and rewrite the letter in own way supplying correct forms based on understanding/interpretation of local semantic context. 6 Seating chart task Tenses, speaker knowledge Free production; grammar task Working in groups, students try to seat people of different and sometimes conflicting backgrounds around one dining table using conditionals. The conditions listed could not be satisfied in their entirety, so students had to try to satisfy as many conditions as they possibly could. View Large Table 3: Pedagogic tasks used in the study Time Task order and name Knowledge/skills addressed through task Foci Gist of task WEEK 1 1 Sentence strips Sequence of tenses between clauses, time markers Controlled production, comprehension; grammar task In groups of 2–3, students discuss meaning of sentence strips representing two parts of conditionals (with blanks for verb tenses). Then students match if -clauses with corresponding main clauses, supplying the correct verb tenses in the blanks and providing an appropriate rationale for their choice. 2 Understanding background knowledge Speaker knowledge Consciousness-raising, comprehension; grammar task Students are given four conditional sentences and asked to explain background knowledge that had to precede formation of each conditional, tying the rationale to form (tense choices). WEEK 2 3 Hedging in data commentaries Tenses and context knowledge Controlled production, comprehension; grammar task Adapted from Swales and Feak (2004) , this task elicited students’ written data commentary of available statistical information within provided table. Students had to frame their data commentary as hypotheses and predictions using conditionals. 4 Global warming causes Tenses; speaker and context knowledge Comprehension, controlled production; grammar task Students discuss causes and effects of global warming using conditionals. For content knowledge, students relied on materials and class discussions from earlier in the semester. WEEK 3 5 Text repair: break-up letter Tenses, context knowledge, time markers Consciousness-raising, controlled production; communicative task Students identify ill-formed conditionals in a break-up letter and rewrite the letter in own way supplying correct forms based on understanding/interpretation of local semantic context. 6 Seating chart task Tenses, speaker knowledge Free production; grammar task Working in groups, students try to seat people of different and sometimes conflicting backgrounds around one dining table using conditionals. The conditions listed could not be satisfied in their entirety, so students had to try to satisfy as many conditions as they possibly could. Time Task order and name Knowledge/skills addressed through task Foci Gist of task WEEK 1 1 Sentence strips Sequence of tenses between clauses, time markers Controlled production, comprehension; grammar task In groups of 2–3, students discuss meaning of sentence strips representing two parts of conditionals (with blanks for verb tenses). Then students match if -clauses with corresponding main clauses, supplying the correct verb tenses in the blanks and providing an appropriate rationale for their choice. 2 Understanding background knowledge Speaker knowledge Consciousness-raising, comprehension; grammar task Students are given four conditional sentences and asked to explain background knowledge that had to precede formation of each conditional, tying the rationale to form (tense choices). WEEK 2 3 Hedging in data commentaries Tenses and context knowledge Controlled production, comprehension; grammar task Adapted from Swales and Feak (2004) , this task elicited students’ written data commentary of available statistical information within provided table. Students had to frame their data commentary as hypotheses and predictions using conditionals. 4 Global warming causes Tenses; speaker and context knowledge Comprehension, controlled production; grammar task Students discuss causes and effects of global warming using conditionals. For content knowledge, students relied on materials and class discussions from earlier in the semester. WEEK 3 5 Text repair: break-up letter Tenses, context knowledge, time markers Consciousness-raising, controlled production; communicative task Students identify ill-formed conditionals in a break-up letter and rewrite the letter in own way supplying correct forms based on understanding/interpretation of local semantic context. 6 Seating chart task Tenses, speaker knowledge Free production; grammar task Working in groups, students try to seat people of different and sometimes conflicting backgrounds around one dining table using conditionals. The conditions listed could not be satisfied in their entirety, so students had to try to satisfy as many conditions as they possibly could. View Large Table 4: Descriptive statistics by testing task Pretest means Post-test means Delayed post-test means Testing task Cog Task Control Cog Task Control Cog Task Control Controlled production 13.59 12.61 13.18 22.06 18.11 14.82 21.29 16.56 14.32 Free production 3.82 4.11 3.55 6.24 5.56 3.77 6.29 5.61 4.45 Comprehension: pictures 2.65 2.83 2.91 3.24 2.94 2.77 2.71 2.50 2.50 Comprehension: sentences 5.12 4.89 3.95 6.65 4.94 4.23 6.24 5.11 4.14 Pretest means Post-test means Delayed post-test means Testing task Cog Task Control Cog Task Control Cog Task Control Controlled production 13.59 12.61 13.18 22.06 18.11 14.82 21.29 16.56 14.32 Free production 3.82 4.11 3.55 6.24 5.56 3.77 6.29 5.61 4.45 Comprehension: pictures 2.65 2.83 2.91 3.24 2.94 2.77 2.71 2.50 2.50 Comprehension: sentences 5.12 4.89 3.95 6.65 4.94 4.23 6.24 5.11 4.14 View Large Table 4: Descriptive statistics by testing task Pretest means Post-test means Delayed post-test means Testing task Cog Task Control Cog Task Control Cog Task Control Controlled production 13.59 12.61 13.18 22.06 18.11 14.82 21.29 16.56 14.32 Free production 3.82 4.11 3.55 6.24 5.56 3.77 6.29 5.61 4.45 Comprehension: pictures 2.65 2.83 2.91 3.24 2.94 2.77 2.71 2.50 2.50 Comprehension: sentences 5.12 4.89 3.95 6.65 4.94 4.23 6.24 5.11 4.14 Pretest means Post-test means Delayed post-test means Testing task Cog Task Control Cog Task Control Cog Task Control Controlled production 13.59 12.61 13.18 22.06 18.11 14.82 21.29 16.56 14.32 Free production 3.82 4.11 3.55 6.24 5.56 3.77 6.29 5.61 4.45 Comprehension: pictures 2.65 2.83 2.91 3.24 2.94 2.77 2.71 2.50 2.50 Comprehension: sentences 5.12 4.89 3.95 6.65 4.94 4.23 6.24 5.11 4.14 View Large Table 5: Descriptive statistics of mean test scores across three groups Score Group Mean SD N Pretest score Cognitive 25.18 5.27 17 Task 24.44 4.33 18 Control 23.59 4.83 22 Post-test score Cognitive 38.06 3.25 17 Task 31.67 5.35 18 Control 25.59 4.85 22 Delayed post-test score Cognitive 36.53 3 17 Task 29.78 4.44 18 Control 25.45 5.71 22 Score Group Mean SD N Pretest score Cognitive 25.18 5.27 17 Task 24.44 4.33 18 Control 23.59 4.83 22 Post-test score Cognitive 38.06 3.25 17 Task 31.67 5.35 18 Control 25.59 4.85 22 Delayed post-test score Cognitive 36.53 3 17 Task 29.78 4.44 18 Control 25.45 5.71 22 View Large Table 5: Descriptive statistics of mean test scores across three groups Score Group Mean SD N Pretest score Cognitive 25.18 5.27 17 Task 24.44 4.33 18 Control 23.59 4.83 22 Post-test score Cognitive 38.06 3.25 17 Task 31.67 5.35 18 Control 25.59 4.85 22 Delayed post-test score Cognitive 36.53 3 17 Task 29.78 4.44 18 Control 25.45 5.71 22 Score Group Mean SD N Pretest score Cognitive 25.18 5.27 17 Task 24.44 4.33 18 Control 23.59 4.83 22 Post-test score Cognitive 38.06 3.25 17 Task 31.67 5.35 18 Control 25.59 4.85 22 Delayed post-test score Cognitive 36.53 3 17 Task 29.78 4.44 18 Control 25.45 5.71 22 View Large Table 5 shows that mean pretest scores were quite similar for all three groups, with 25.18, 24.44, and 23.59 for cognitive, task-supported, and control group, respectively, the average standard deviation (SD) among all three groups being 4.778. On the post-test, average score of the cognitive group participants was 38.06, the one for the task-supported group participants was 31.67, and the one for the control group participants was 25.59, the SD for the entire sample being 6.871. Thus, the average score of cognitive group participants expanded by 13 points, that of task-supported group participants grew by 7 points, while that of control group increased by 2 points. The delayed post-test scores were 36.53, 29.78, and 25.45 for the cognitive, task-supported, and control groups, respectively, which demonstrate an average loss of about 2 points for both treatment groups; the score of control group stayed approximately the same, declining only by an average of one tenth of a point between the post-test and the delayed post-test. ANOVA comparison for all three groups In the present study, a repeated-measures within-subjects ANOVA was chosen for the main statistical procedure because the same subjects contributed to test means at different points of the experiment. Mauchly’s test of sphericity was conducted to ensure that the variances of the differences between conditions were equal. The p -value obtained on this test (.448) was not significant; therefore, the condition of sphericity could be considered satisfied for the present data and study context. Normality assumptions (skewness and kurtosis) were also met for this data set. The data in Table 6 below provide an overview of tests of within-subjects effects, while Table 7 reports the results of multivariate tests. Table 6: Repeated measures ANOVA: Tests of within-subjects effects Source df Mean square F Sig. ** Time Sphericity Assumed 2 881.486 70.66 .000 Greenhouse-Geisser 1.942 907.824 70.66 .000 Huynh-Feldt 2.000 881.486 70.66 .000 Lower-bound 1.000 1762.973 70.66 .000 Time * GROUP Sphericity Assumed 4 170.188 13.64 .000 Greenhouse-Geisser 3.884 175.273 13.64 .000 Huynh-Feldt 4.000 170.188 13.64 .000 Lower-bound 2.000 340.376 13.64 .000 Source df Mean square F Sig. ** Time Sphericity Assumed 2 881.486 70.66 .000 Greenhouse-Geisser 1.942 907.824 70.66 .000 Huynh-Feldt 2.000 881.486 70.66 .000 Lower-bound 1.000 1762.973 70.66 .000 Time * GROUP Sphericity Assumed 4 170.188 13.64 .000 Greenhouse-Geisser 3.884 175.273 13.64 .000 Huynh-Feldt 4.000 170.188 13.64 .000 Lower-bound 2.000 340.376 13.64 .000 ** p < .01. View Large Table 6: Repeated measures ANOVA: Tests of within-subjects effects Source df Mean square F Sig. ** Time Sphericity Assumed 2 881.486 70.66 .000 Greenhouse-Geisser 1.942 907.824 70.66 .000 Huynh-Feldt 2.000 881.486 70.66 .000 Lower-bound 1.000 1762.973 70.66 .000 Time * GROUP Sphericity Assumed 4 170.188 13.64 .000 Greenhouse-Geisser 3.884 175.273 13.64 .000 Huynh-Feldt 4.000 170.188 13.64 .000 Lower-bound 2.000 340.376 13.64 .000 Source df Mean square F Sig. ** Time Sphericity Assumed 2 881.486 70.66 .000 Greenhouse-Geisser 1.942 907.824 70.66 .000 Huynh-Feldt 2.000 881.486 70.66 .000 Lower-bound 1.000 1762.973 70.66 .000 Time * GROUP Sphericity Assumed 4 170.188 13.64 .000 Greenhouse-Geisser 3.884 175.273 13.64 .000 Huynh-Feldt 4.000 170.188 13.64 .000 Lower-bound 2.000 340.376 13.64 .000 ** p < .01. View Large Table 7: Repeated-measures ANOVA: Multivariate tests for performance across time in three groups Effect Value F Sig. Partial Eta Squared Time Pillai's Trace 0.74 77.08 .000 0.74 Wilks' Lambda 0.26 77.08 .000 0.74 Hotelling's Trace 2.91 77.08 .000 0.74 Roy's Largest Root 2.91 77.08 .000 0.74 Time * GROUP Pillai's Trace 0.54 9.86 .000 0.27 Wilks' Lambda 0.47 12.12 .000 0.31 Hotelling's Trace 1.11 14.44 .000 0.36 Roy's Largest Root 1.10 29.7 .000 0.52 Effect Value F Sig. Partial Eta Squared Time Pillai's Trace 0.74 77.08 .000 0.74 Wilks' Lambda 0.26 77.08 .000 0.74 Hotelling's Trace 2.91 77.08 .000 0.74 Roy's Largest Root 2.91 77.08 .000 0.74 Time * GROUP Pillai's Trace 0.54 9.86 .000 0.27 Wilks' Lambda 0.47 12.12 .000 0.31 Hotelling's Trace 1.11 14.44 .000 0.36 Roy's Largest Root 1.10 29.7 .000 0.52 View Large Table 7: Repeated-measures ANOVA: Multivariate tests for performance across time in three groups Effect Value F Sig. Partial Eta Squared Time Pillai's Trace 0.74 77.08 .000 0.74 Wilks' Lambda 0.26 77.08 .000 0.74 Hotelling's Trace 2.91 77.08 .000 0.74 Roy's Largest Root 2.91 77.08 .000 0.74 Time * GROUP Pillai's Trace 0.54 9.86 .000 0.27 Wilks' Lambda 0.47 12.12 .000 0.31 Hotelling's Trace 1.11 14.44 .000 0.36 Roy's Largest Root 1.10 29.7 .000 0.52 Effect Value F Sig. Partial Eta Squared Time Pillai's Trace 0.74 77.08 .000 0.74 Wilks' Lambda 0.26 77.08 .000 0.74 Hotelling's Trace 2.91 77.08 .000 0.74 Roy's Largest Root 2.91 77.08 .000 0.74 Time * GROUP Pillai's Trace 0.54 9.86 .000 0.27 Wilks' Lambda 0.47 12.12 .000 0.31 Hotelling's Trace 1.11 14.44 .000 0.36 Roy's Largest Root 1.10 29.7 .000 0.52 View Large Table 6 shows that the p value is <.001, demonstrating that the differences between the subjects’ mean scores on all three tests were statistically significant for each group pairing (i.e. between cognitive and task-supported groups, between task-supported and control groups, and between cognitive and control groups). As can be seen from Table 7 , multivariate tests focusing on the interaction between time and group variables demonstrate medium-strength effect sizes (partial eta squared). To check the specific areas of interaction between groups, Scheffe’s post hoc comparison was conducted; the results are represented in Table 8 below. Table 8: Repeated measures ANOVA: Multiple comparisons, Scheffe’s post hoc test (I) Group (J) Group Mean difference (I-J) Standard Error Sig. Cognitive Task 4.63 * 1.251 .002 Control 8.38 * 1.195 .000 Task Cognitive −4.63 * 1.251 .002 Control 3.75 * 1.176 .009 Control Cognitive −8.38 * 1.195 .000 Task −3.75 * 1.176 .009 (I) Group (J) Group Mean difference (I-J) Standard Error Sig. Cognitive Task 4.63 * 1.251 .002 Control 8.38 * 1.195 .000 Task Cognitive −4.63 * 1.251 .002 Control 3.75 * 1.176 .009 Control Cognitive −8.38 * 1.195 .000 Task −3.75 * 1.176 .009 * p < .01. View Large Table 8: Repeated measures ANOVA: Multiple comparisons, Scheffe’s post hoc test (I) Group (J) Group Mean difference (I-J) Standard Error Sig. Cognitive Task 4.63 * 1.251 .002 Control 8.38 * 1.195 .000 Task Cognitive −4.63 * 1.251 .002 Control 3.75 * 1.176 .009 Control Cognitive −8.38 * 1.195 .000 Task −3.75 * 1.176 .009 (I) Group (J) Group Mean difference (I-J) Standard Error Sig. Cognitive Task 4.63 * 1.251 .002 Control 8.38 * 1.195 .000 Task Cognitive −4.63 * 1.251 .002 Control 3.75 * 1.176 .009 Control Cognitive −8.38 * 1.195 .000 Task −3.75 * 1.176 .009 * p < .01. View Large Results presented in Table 8 suggest that the relationships between all group pairings were statistically significant with consistently observable p < .01. A visual representation of the how participants of each group scored across all three tests is given in Figure 1 below. Figure 1: View largeDownload slide Means of test performance over time Figure 1: View largeDownload slide Means of test performance over time Furthermore, ANOVAs to compare performance on each testing task were produced to determine the areas of greatest and least improvement over the period of the study. ANOVA for controlled production. Overall differences for controlled production were found to be statistically significant (between-subjects effects for interaction between time and group: F = 8.4, p = .001, partial eta-squared = 0.237). To avoid possible redundancy in discussion, only the data demonstrating actual interactions between groups will be reported. Scheffe’s post hoc test showed that the differences between the cognitive and task-supported groups and the cognitive and control groups were statistically significant, while the differences between the task-supported and the control groups were not. ANOVA for free production. For the free production part of the tests (between-subjects effects for interaction between time and group: F = 3.5, p = .01, partial eta squared = 0.11), the differences between the cognitive and the control groups, as well as between the task-supported and the control groups, were found to be statistically significant, while the difference between the cognitive and the task-supported groups was not. ANOVA for ‘comprehension: pictures’. For the ‘comprehension: pictures’ part, none of the interactions between groups proved to be statistically significant (between-subjects effects for interaction between time and group: F = 1.0, p = .4, partial eta squared = 0.03). The lack of significance may be related to the small number of items in this part of the test (four total). ANOVA for ‘comprehension: sentences’. For the final part of the test, ‘comprehension: sentences’, the differences between all group pairings proved to be statistically significant, albeit with marginal effect size ( F = 3.4, p = .04, partial eta squared = 0.06). Accordingly, these findings suggest that production parts of the tests—in particular, the controlled production—were the areas of biggest improvement over the course of the study. t -test comparison between cognitive and task-supported groups on overall test gain scores The descriptive statistics pertaining to the overall test scores and the production scores indicate clear differences between post-test outcomes in cognitive as opposed to task-supported groups. Descriptive statistics for the overall gains scores between the cognitive and task-supported groups are presented in Table 9 below. Table 9: Descriptive statistics for overall gains scores: Cognitive vs. task-supported groups Group N Mean SD SE mean Cognitive 17 12.82 5.01 1.21 Task-supported 18 7.28 5.43 1.28 Group N Mean SD SE mean Cognitive 17 12.82 5.01 1.21 Task-supported 18 7.28 5.43 1.28 View Large Table 9: Descriptive statistics for overall gains scores: Cognitive vs. task-supported groups Group N Mean SD SE mean Cognitive 17 12.82 5.01 1.21 Task-supported 18 7.28 5.43 1.28 Group N Mean SD SE mean Cognitive 17 12.82 5.01 1.21 Task-supported 18 7.28 5.43 1.28 View Large As can be seen from the table, the mean overall gain score for the cognitive group was 12.82, while the mean overall gain score for the task-supported group was 7.28, with the SD being in the low 5+ range (SE = 1.216 and 1.280 for the cognitive and the task-supported groups, respectively). The overview of the t -test comparison between these trends is provided in Table 10 . Table 10: T-test on overall gains scores: Cognitive and task-supported groups t df Sig. (2-tailed) Mean difference 3.13 33 .004 5.55 t df Sig. (2-tailed) Mean difference 3.13 33 .004 5.55 View Large Table 10: T-test on overall gains scores: Cognitive and task-supported groups t df Sig. (2-tailed) Mean difference 3.13 33 .004 5.55 t df Sig. (2-tailed) Mean difference 3.13 33 .004 5.55 View Large An independent samples t -test procedure was chosen because differences between two distinct groups were being measured through comparing individual/independent scores on two sets of tests. Levene’s test for equality of variances ( p =.826; F = 0.049) confirmed that equal variances can be assumed for this analysis. The two-tailed test for equality of means proved to be statistically significant with p = .004 with t (33) = 3.133. Summary of results The cognitive group obtained greater production score and overall test score gains between the pretest and the post-test, than both the task-supported and the control groups did. Participants of the cognitive and the task-supported groups were able to obtain greater test score gains between the pretest and the post-test, than the participants of the control group, thus providing support for the value of task-based classroom instruction for the acquisition of English conditionals. DISCUSSION Meaning of results This section will address answers to the research questions and general implications of the findings of this study. Research Question 1 : Does task-supported instruction produce an effect for L2 development of conditional constructions? Research Question 2 : Does task-supported instruction with added CL insights produce an effect for L2 development of conditional constructions? Research Question 3 : Does the addition of CL insights into task-supported instruction produce greater L2 development of conditional constructions than task-supported instruction alone? In response to the first research question, the initially assumed null hypothesis can be rejected, as the task-supported group outperformed the control group on both the post-test and the delayed post-test: the differences between the scores of these two groups were statistically significant. The second null hypothesis (research question 2) can also be rejected: statistical analyses demonstrated that the cognitive group outperformed the control group on both the post-test and the delayed post-test. In sum, both treatment groups outperformed the control group, thus demonstrating that explicit instruction of conditionals supported by pedagogic tasks is more effective for L2 development of this target form than no instruction. Furthermore, in response to the third question, the statistical analyses demonstrated that the cognitive group outperformed the task-supported group on both the post-test and the delayed post-test. In addition to general trends regarding effectiveness of instruction types, the analysis of post-test data revealed the following trends worth highlighting and discussing in greater detail: Changes in comprehension scores . Prior research has generally assumed that comprehension knowledge is supposed to precede production and, accordingly, the receptive knowledge generally exceeds the productive knowledge for L2 learners ( Lightbown 1985 , 2000 ; Ellis 2004 ). Judging by pretest scores on comprehension sections, all groups initially were at approximately the same level. After the instructional treatment, almost all of the cognitive group participants made consistent, albeit small (around 1–2 points), gains on the comprehension sections, while for participants of the task-supported and control groups, the gains in comprehension scores were not as consistent. Even though the numbers of comprehension items in the tests were too small for inferential statistics, this small-scale trend is suggestive of improved comprehension of conditionals that resulted from exposure to the CL treatment. Changes in production scores . Production parts of the tests were where most improvement took place between the pretest and the post-test, suggesting that the instructional treatment was instrumental for promoting production. It can be inferred that pedagogical tasks and the teacher-facilitated instructional PowerPoint presentations were effective methods for improving the participants’ production skills. Between the two treatment groups, the cognitive group outperformed the task-supported group on the production parts of the test, and the difference was calculated to be statistically significant with p < .01. The implication of this finding is that the combination of cognitive and task-supported treatment was more effective in improving the participants’ production of conditionals than the task-supported treatment alone. Conducting further research focusing specifically on production aspects of conditional phrases would be necessary, before any further, more certain conclusions could be drawn in this regard. Changes in overall test gain scores . The ANOVA and accompanying comparisons measuring gains between tests demonstrated that the difference between the gains made by participants of the two treatment groups was statistically significant with p < .001. This statistical finding suggests that the combination of the CL presentation of conditionals complemented by pedagogical tasks was generally more effective than the task-supported treatment alone: the gains made by the cognitive group participants were almost double compared to those of the task-supported group participants. It would be fair to hypothesize that the presentation of conditionals to the cognitive group was more cohesive and meaningful because it had a consistent foundation in the shape of CL presentation of language. Accordingly, the performance of the cognitive group may be attributed to the presence of the general focus on meaning and the CL insights emphasizing the role of local context and speaker’s background knowledge when creating a conditional phrase. The benefits associated with focus on meaning were also highlighted by the CL group subjects participating in retrospective interviews (for a full discussion of qualitative data pertaining to this study, see Jacobsen (2015) . Furthermore, since both treatment groups demonstrated improvement on test scores between the pretest and the post-test and since tasks were the instructional element shared by both groups, it is highly likely that tasks were the crucial instructional aspect contributing to L2 development within both treatment groups in comparison with the control group. Summative findings from the retrospective interview data demonstrated that students from both groups explicitly favored tasks that were relatively more complex (i.e. some type of cognitive effort was necessary for their completion), included relatively richer contextual knowledge, and required production and group work. While this study did not explicitly test for effects of task complexity on L2 development, this finding appears to be consistent with Robinson’s (2005 , 2011 ) research on cognitive task demands and their effects upon L2 development. Taking these trends into a broader theoretical context, it is important to highlight the fact that focus on form alone (as received by the task-supported group) was not demonstrated to be as effective as the focus on form combined with CL insights. While the support for the instruction (consciousness-raising tasks) mattered and produced gains for both treatment groups, the underlying theory of language assumed in teacher-facilitated presentation played a further facilitative role for L2 development. It is possible that tasks helped further contextualize the meaning conveyed through the cognitive explanation, and thus the use of tasks may have further enhanced the performance of the cognitive group. However, since we did not have a cognitive-only (no tasks) group in this study, it is not possible to specify the precise nature of interaction between the two variables. (Findings discussed in Jacobsen (2015) also supported the use of pedagogic tasks for cognitive-informed and traditional instruction.) Since the cognitive group outperformed the task-supported group, it can be implied that the CL theory has the potential of further facilitating successful language teaching. General implications of the study This study aimed to test the applicability of CL theory to the instruction of conditionals within a mixed-methods design, using tasks as pedagogical means supporting the delivery of instructional content. Both treatment conditions—cognitive and task-supported—were effective for producing L2 development of the target form, as measured by the tests. In line with findings of DeKeyser and Prieto Botana’s (2015) meta-analysis supporting the benefit of explicit instruction for structures of greater complexity, this study demonstrated that explicit focus on form was effective for the classroom instruction of conditionals. It is important to discuss the interaction between CL instructional approach and raising metalinguistic awareness. Recent research by Roehr-Brackin (2014) suggests that explicit knowledge and learner’s reliance on metalinguistic awareness strategies can be a powerful tool for promoting L2 learning and use. If adapted to learners’ level, CL provides a framework for raising metalinguistic awareness systematically, which is a clear benefit of applying CLs in L2 instruction. Furthermore, the combination of the cognitive and task-supported instruction was more effective than the task-supported instruction alone. Pedagogic tasks proved to be an appropriate method for exploring authentic usage contexts. CLs appears to provide a better theory of language in general than the traditional (formal and structural) approaches to language analysis. Being a usage-based theory of language, CLs is highly compatible with research on task-based teaching, which also emphasizes language learning through meaningful content and contextualized language use. Task-supported instruction can function as a pedagogical platform complementary to the process of adapting the CL theory to language classroom. Combined with qualitative findings (see Jacobsen 2015 ), these results suggest that the focus on usage-based patterns and meaning indeed made a difference in the subjects’ performance. While pedagogic tasks are a valuable addition to the instructional process, the improved teaching methods alone are not sufficient. A more accurate and insightful theory of language is needed to make language teaching more meaningful overall. Limitations and future directions The study has a number of limitations. First of all, with N = 57, the sample size was relatively small, preventing the results from becoming more generalizable. It is also possible that some of the findings can be attributed to the context of this particular graduate EAP program as well as to the educational and cultural backgrounds of the majority participants. Replication with students from other backgrounds (i.e. taking into account other prior instructional experiences) would be instrumental for establishing the degree of generalizability of this study’s findings. Furthermore, echoing previously documented limitations ( Boers 2013 ), it was not possible to account for learners’ individual differences; additional data on learners’ motivation, working memory, and general aptitude (as defined by specific contexts) would need to be collected to determine the effects of possible intervening variables. Also, the whole instructional treatment lasted three weeks, but given that CL methods are relatively novel in classroom instruction, it would be necessary to test the long-term effects of a similar treatment. Establishing more specific effects of longitudinal and systematic classroom exposure to CL theory had previously been proposed as an area of recommended inquiry by other researchers ( Holme 2012 ; Tyler 2012 ). Furthermore, teacher buy-in and appropriate training would be crucial for successful replication of applied CL research. Another limitation of this study is such that the CL materials can only be used by individuals who both received prior training in CL and believe in its usefulness as a paradigm of language teaching. In a way, promoting CL as a basis for an L2 instructional system would require significant popularization of CL in teacher training departments and a CL-oriented shift in language teacher training curriculum. Until CL becomes a mainstream part of language teaching, we could not adopt directional hypotheses for L2 studies involving CL-informed methodology. In general, while this study does provide support for using adapted CL theory in the classroom, the research conducted thus far is not sufficient for establishing CL theory’s superiority over traditional instruction; additional future research needs to examine how CL theory in combination with systematic pedagogical interventions can be used for teaching other grammatical structures. Notes Natalia Dolgova Jacobsen is Teaching Assistant Professor of English for Academic Purposes at the George Washington University in Washington, DC. Her research interests include applied cognitive and corpus linguistics, task-based language teaching, second language writing in professional contexts, and English for Academic/Specific Purposes. Address for correspondence : English for Academic Purposes (EAP) Program, George Washington University, 2114 G Street NW, Washington, DC 20052, USA. < natalia@gwu.edu > 1 This study uses the task-supported model because of local context constraints (see ‘Methods’ section for details); however, TBLT is at the ideological root of task-supported instruction and is therefore addressed in this section. 2 The term ‘predictive’ refers to any expression of prediction conditioned ‘upon some unrealized event’ ( Dancygier and Sweetser 2005 : 28). If we use the traditional ESL terminology, all three conditional types—factual, hypothetical, and counterfactual—were addressed in the study. References Achard M. 2004 . ‘Grammatical instruction in the natural approach: A cognitive grammar view’ in Achard M. Niemeier S. (eds): Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching . Mouton de Gruyter . Achard M. Niemeier S. (eds). 2004 . Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching . Mouton de Gruyter . Boers F. 2013 . ‘Cognitive Semantic ways of teaching figurative phrases. An assessment’ in Gonzálvez-García F. Peña Cervel M. S. Pérez Hernández L. (eds): Metaphor and Metonymy Revisited Beyond the Contemporary Theory of Metaphor . John Benjamins Boers F. Lindstromberg S. . 2008 . Cognitive Linguistic Approaches to Teaching Vocabulary and Phraseology . Mouton de Gruyter . Bygate M. Skehan P. Swain M. (eds). 2001 . Researching Pedagogic Tasks, Second Language Learning, Teaching and Testing . Longman . Cadierno T. 2008 . ‘ Learning to talk about motion ’ in Robinson P. Ellis N. (eds): Handbook of Cognitive Linguistics and Second Language Acquisition . Routledge . Cadierno T. Robinson P. . 2009 . ‘Language typology, task complexity and the development of L2 lexicalization patterns for describing motion events ,’ AnnualReview of Cognitive Linguistics 6 : 245 – 77 . Google Scholar CrossRef Search ADS Celce-Murcia M. Larsen-Freeman D. . 1999 . The Grammar Book: An ESL/EFL Teacher’s Course . Heinle & Heinle . Comrie B. 1986 . ‘Conditionals: a typology’ in Traugott E. ter Meulen A. Snitzer Reilly J. Ferguson C. (eds): On conditionals . Cambridge University Press . Dancygier B. Sweetser E. . 2005 . Mental Spaces in Grammar: Conditional Constructions . Cambridge University Press . DeKeyser R. Prieto Botana G. . 2015 . ‘ The effectiveness of processing instruction in L2 grammar acquisition: A narrative review ,’ Applied Linguistics 36 : 290 – 305 . Google Scholar CrossRef Search ADS Dirven R. 2001 . ‘English phrasal verbs: theory and didactic application’ in Putz M. Niemeier S. Dirven R. (eds): Applied cognitive linguistics II: Language Pedagogy . Mouton de Gruyter . Eckerth J. 2008a . ‘ Investigating consciousness-raising tasks: Pedagogically-targeted and non-targeted learning gains ,’ International Journal of Applied Linguistics 18 : 119 – 45 . Google Scholar CrossRef Search ADS Eckerth J. 2008b . ‘ Task-based learner-learner interaction ’ in Eckerth J. Siekmann S. (eds): Task-based Language Learning and Teaching: Theoretical, Methodological, and Pedagogical Perspectives . Peter Lang . Ellis N. 2002 . ‘ Frequency effects in language processing: A review with implications for theories of implicit and explicit language acquisition ,’ Studies in Second Language Acquisition 24 : 143 – 88 . Ellis R. 2003 . Task-based Language Learning and Teaching . Oxford University Press . Ellis R. 2004 . ‘ The definition and measurement of L2 explicit knowledge ,’ Language Learning 54 : 227 – 75 . Google Scholar CrossRef Search ADS Ellis R. 2009 . ‘ Task-based language teaching: Sorting out the misunderstandings ,’ International Journal of Applied Linguistics 19 : 222 – 46 . Google Scholar CrossRef Search ADS Evans J. Over D. . 2004 . If . Oxford University Press . Fauconnier G. Turner M. . 2002 . The Way We Think: Conceptual Blending and the Mind's Hidden Complexities . Basic Books . Fordyce K. 2014 . ‘ The differential effects of explicit and implicit instruction on EFL learners’ use of epistemic stance ,’ Applied Linguistics 35 : 6 – 28 . Google Scholar CrossRef Search ADS Fotos S. 1993 . ‘ Consciousness raising and noticing through focus on form: Grammar task performance versus formal instruction ,’ Applied Linguistics 14 : 385 – 407 . Google Scholar CrossRef Search ADS Fotos S. 1994 . ‘ Integrating grammar instruction and communicative language use through grammar consciousness-raising tasks ,’ TESOL Quarterly 28 : 323 – 51 . Google Scholar CrossRef Search ADS Fotos S. Ellis R. . 1991 . ‘ Communicating about grammar: A task-based approach ,’ TESOL Quarterly 25 : 605 – 28 . Google Scholar CrossRef Search ADS De la Fuente M. 2006 . ‘ Classroom L2 vocabulary acquisition: Investigating the role of pedagogical tasks and form-focused instruction ,’ Language Teaching Research 10 : 263 – 95 . Google Scholar CrossRef Search ADS Han Y. Ellis R. . 1998 . ‘ Implicit knowledge, explicit knowledge and general language proficiency ,’ Language Teaching Research 2 : 1 – 23 . Google Scholar CrossRef Search ADS Holme R. 2009 . Cognitive Linguistics and Language Teaching . Palgrave MacMillan . Holme R. 2012 . ‘ Cognitive linguistics and the second language classroom ,’ TESOL Quarterly 46 : 6 – 29 . Google Scholar CrossRef Search ADS Hulstijn J. H. de Graaff R. . 1994 . ‘ Under what conditions does explicit knowledge of a second language facilitate the acquisition of implicit knowledge? A research proposal ,’ AILA Review 11 : 97 – 112 . Jackson F. 1991 . Conditionals . Oxford University Press . Jacobsen N. D. 2012 . Applying Cognitive Linguistics and Task-supported Language Teaching to Instruction of English Conditional Phrases . Washington, DC: Georgetown University , Unpublished Dissertation. Jacobsen N. 2015 . ‘A cognitive linguistic analysis of English conditionals in English for Academic Purposes (EAP) instruction: Implications from sociocultural theory’ in Masuda K. Arnett C. Labarca A. (eds): Cognitive Linguistics and Sociocultural Theory: Applications for Second and Foreign Language Teaching . Mouton de Gruyter . Langacker R. W. 2008 . ‘Cognitive grammar as a basis for language instruction’ in Robinson P. Ellis N. (eds): Handbook of Cognitive Linguistics and Second Language Acquisition . Routledge . Lightbown P. 1985 . ‘ Great expectations: Second language acquisition research and classroom teaching ,’ Applied Linguistics 6 : 173 – 89 . Google Scholar CrossRef Search ADS Lightbown P. 2000 . ‘ Anniversary article: Classroom SLA research and second language teaching ,’ Applied Linguistics 21 : 431 – 62 . Google Scholar CrossRef Search ADS Liu D. 2010 . ‘ Going beyond patterns: Involving cognitive analysis in the learning of collocations ,’ TESOL Quarterly 44 : 4 – 30 . Google Scholar CrossRef Search ADS Lindstromberg S. 1996 . ‘ Prepositions: meaning and method ,’ ELT Journal 50 : 225 – 36 . Google Scholar CrossRef Search ADS Lindstromberg S. Boers F. . 2005 . ‘ From movement to metaphor with manner-of-movement verbs ,’ Applied Linguistics 26 : 241 – 61 . Google Scholar CrossRef Search ADS Littlemore J. Low G. . 2006 . ‘ Metaphoric competence, second language learning, and communicative language ability ,’ Applied Linguistics 27 : 268 – 94 . Google Scholar CrossRef Search ADS Long M. 2000 . ‘Focus on form in task-based language teaching’ in Lambert R. Shohamy E. (eds): Language Policy and Pedagogy: Essays in Honor of A. Ronald Walton . John Benjamins . Meunier F. 2008 . ‘Corpora, cognition and pedagogical grammars’ in de Knop S. D. de Rycker T. D. (eds): Cognitive Approaches to Pedagogical Grammar: A Volume In honour of Rene Dirven . Mouton de Gruyter . Moder C. L. 2010 . ‘To learn effectively: Grammatical constructions for second language grammar instruction’. Paper presented at AAAL, Atlanta, GA. Niemeier S. Reif M. . 2008 . ‘Applying cognitive grammar to tense-aspect teaching’ in de Knop S. de Rycker T. (eds): Cognitive Approaches to Pedagogical Grammar: A Volume In honour of Rene Dirven . Mouton de Gruyter . Norris J. Ortega L. . 2000 . ‘ Effectiveness of L2 instruction: A research synthesis and quantitative meta-analysis ,’ Language Learning 50 : 417 – 528 . Google Scholar CrossRef Search ADS Révész A. 2009 . ‘ Task complexity, focus on form, and second language development ,’ Studies in Second Language Acquisition 31 :/ 3 : 437 – 470 . Google Scholar CrossRef Search ADS Robinson P. 2001 . ‘Task complexity, cognitive resources, and syllabus design’ in Robinson P. (ed.): Cognition and Second Language Instruction . Cambridge University Press . Robinson P. 2005 . ‘ Cognitive complexity and task sequencing ,’ International Review of Applied Linguistics 43 : 1 – 32 . Google Scholar CrossRef Search ADS Robinson P. 2011 . ‘ Task-based language learning: A review of issues ,’ Language Learning 61(Supplement 1) : 1 – 36 . Google Scholar CrossRef Search ADS Robinson P. Ellis N. (eds.). 2008 . Handbook of Cognitive Linguistics and Second Language Acquisition . Routledge . Roehr-Brackin K. 2014 . ‘ Explicit knowledge and processes from a usage-based perspective: The developmental trajectory of an instructed L2 learner ,’ Language Learning 64 : 771 – 808 . Google Scholar CrossRef Search ADS Schmidt R. 2001 . ‘Attention’ in Robinson P. (ed.): Cognition and Second Language Instruction . Cambridge University Press . Samuda V. Bygate M. . 2008 . Tasks in Second Language Learning . Palgrave Macmillan . Sharwood Smith M. 1981 . ‘ Consciousness-raising and the second language learner ,’ Applied Linguistics 2 / 2 : 159 – 68 . Google Scholar CrossRef Search ADS Sharwood Smith M. 1993 . ‘ Input enhancement in instructed SLA: Theoretical bases ,’ Studies in Second Language Acquisition 15 : 165 – 79 . Google Scholar CrossRef Search ADS Sharwood Smith M. 2007 . ‘Morphological and syntactic awareness in foreign/second language learning’ in Cenoz J. Hornberger N. (eds): Encyclopedia of Language and Education . Kluwer Academic Publishers . Spada N. Tomita Y. . 2010 . ‘ Interactions between type of instruction and type of language feature: A meta-analysis ,’ Language Learning 60 : 263 – 308 . Google Scholar CrossRef Search ADS Swales J. Feak C. . 1998 . Academic Writing for Graduate Students. University of Michigan Press . Taylor J. 1997 . ‘Conditionals and polarity’ in Athanasiadou A. Dirven R. (eds): On Conditionals Again . John Benjamins . Tyler A. 2012 . Cognitive Linguistics and Second Language Learning: Theoretical Basics and Experimental Evidence . Routledge Tyler A. Evans V. . 2004 . ‘ Applying cognitive linguistics to pedagogical grammar: The case of over ’ in Achard M. Niemeier S. (eds): Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching . Mouton de Gruyter . Tyler A. Mueller C. Ho V. . 2011 . ‘ Applying cognitive linguistics to learning the semantics of English to, for, and at: An experimental investigation ,’ Vigo International Journal of Applied Linguistics 8 : 181 – 205 . Van den Branden K. Bygate M. Norris J. . (eds) 2009 . Task-based Language Teaching: A Reader . John Benjamins . Werth P. 1997 . ‘Conditionality as cognitive distance’ in Athanasiadou A. Dirven R. (eds): On Conditionals Again . John Benjamins . Woods M. 1997 . Conditionals . Clarendon Press . Yasuda S. 2010 . ‘ Learning phrasal verbs through conceptual metaphors: A case of Japanese EFL learners ,’ TESOL Quarterly 44 : 250 – 73 . Google Scholar CrossRef Search ADS © Oxford University Press 2016 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Applied Linguistics Oxford University Press

The Best of Both Worlds: Combining Cognitive Linguistics and Pedagogic Tasks to Teach English Conditionals

Loading next page...
 
/lp/ou_press/the-best-of-both-worlds-combining-cognitive-linguistics-and-pedagogic-hdUyjVeVaF
Publisher
Oxford University Press
Copyright
© Oxford University Press 2016
ISSN
0142-6001
eISSN
1477-450X
D.O.I.
10.1093/applin/amw030
Publisher site
See Article on Publisher Site

Abstract

Abstract Due to their internal complexity, English conditionals represent a challenge for L2 learners; furthermore, most ESL grammar books lack precise explanations of conditional meanings and usage contexts. As a solution, cognitive linguistic research on conditionals ( Dancygier and Sweetser 2005 ), based on theory of conceptual blending and mirroring human cognition processes, could be very beneficial for L2 learners. In this study, elements of cognitive linguistic analysis of English hypothetical conditionals were incorporated into L2 learning materials. The design of the study was quasi-experimental, with three groups ( N = 57) participating: cognitive, task-supported, and control. Pedagogic consciousness-raising tasks were used to provide a context for meaningful practice for cognitive and task-supported groups. The research questions explored the efficacy of different types of instruction for L2 development of English conditional phrases, as measured by a post-test and a delayed post-test. The findings of t -test and analysis of variance procedures ( p = .001 ) indicate that participants from the cognitive group outperformed the participants from the task-supported and control groups and demonstrated a higher degree of understanding of conditional usage contexts. INTRODUCTION The value of using cognitive linguistics for language teaching Cognitive linguistic (CL) theory of language and usage-based approaches to language pedagogy have been gaining recognition by second language acquisition (SLA) researchers ( Robinson and Ellis 2008 ; Tyler 2012 ) over the course of the past couple decades. CLs assumes that language is composed solely of form-meaning pairings, or ‘symbolic units’ ( Langacker 2008 ). Under this analysis, syntax and morphology cannot be strictly separated from lexis, but rather, all linguistic units carry some sort of meaning. Investigating the meaning of symbolic units can be equated with exploring the nature of conceptualization. Ways to conceptualize a specific event or entity, and thus the message, are inherently subjective, as they vary widely depending on the speaker (his/her background knowledge and choice of discourse focus) and the assessment of what the listener already knows. Accordingly, to explore the meaning of a language expression, one would need to determine the conventional linguistic choices, the context of use, and the background knowledge it appears to evoke. CL view of language thus allows to uncover the semantic motivation behind certain syntactic, grammatical, or morphological phenomena. Using select, carefully adapted for unprepared audiences, and contextualized CL concepts in L2 instruction can reveal the perspective of a native speaker and make form-meaning mappings relatively transparent for L2 learners. Accordingly, the underlying conceptual characteristics of CLs make it a good candidate for the role of providing a comprehensive theory that could successfully support L2 instruction ( Achard and Niemeier 2004 ; Tyler and Evans 2004 ; Tyler 2012 ). A number of researchers conducted studies using CL theory to inform English L2 teaching. CL approaches were successfully applied to teaching of phrasal verbs ( Dirven 2001 ; Liu 2010 ), metaphorical/idiomatic language ( Lindstromberg and Boers 2005 ; Littlemore and Low 2006 ; Boers & Lindstromberg. 2008 ; Yasuda 2010 ), and grammar (prepositions: Lindstromberg 1996 ; Tyler and Evans 2004 ; Tyler Mueller and Ho 2011 ; tense and aspect: Niemeier and Reif 2008 ). (For a more detailed overview of these and other studies, see Jacobsen 2012 , 2015 .) Despite the rich qualitative support for adapting CLs to L2 instruction, the existing number of experimental and/or quantitative research studies is not sufficient to provide solid empirical backing to L2 instructional benefits associated with CLs: that is, some of the above-mentioned studies did not employ a rigorous research design or use statistical procedures that would allow for corresponding findings to be considered generalizable. Also, there is a dearth of research into how CL L2 applications can be paired or combined with best practices in language pedagogy. For instance, Achard (2004) examined CL pairing with the Natural Approach, while Holme (2009) explored CL applications of total physical response (TPR) to the L2 classroom. While both of these pedagogical approaches are valid for a limited range of language structures, it is challenging to apply them in all language-teaching contexts. Boers (2013) reached a similar conclusion about the limits of L2 applicability of CL metaphor theory: while concepts from metaphor research have been helpful for teaching figurative language, existing applied CL research does not promote learner recognition and retention of new figurative phrases, nor does it overcome the effects of learners’ individual differences, such as motivation, proficiency, and learning styles. In sum, the generalizability of CL advantages for L2 instruction still appears to be limited, and much more research needs to be conducted to further validate or possibly refine the previously established benefits. The goal of the present study is to investigate the relevance of CL for L2 teaching, addressing the above-mentioned gaps as much as possible. The study focuses on the efficacy of applying CL analysis of English conditionals to L2 instruction, using a mixed-methods design and supplementing CL approach with task-supported language teaching. The concept of conditionals: Why are they problematic for L2 learners? Conditional constructions allow humans to negotiate between several logical scenarios and thus capture various consequences of their actions or relevant circumstances, including making predictions about events that have not yet happened in the present or that did not happen in the past. The surface form of English conditional constructions is frequently referred to as if P, then Q , where P is the conditional clause, and Q is the head clause ( Comrie 1986 ; Jackson 1991 ; Woods 1997 ; Evans and Over 2004 ). The P-clause sets up a ‘theoretical’ situation and marks it as remote from actuality, the Q-clause takes it on to a further outcome, marking it with some degree of probability ( Werth 1997 : 252). The classification most widely used in ESL contexts is provided in Table 1 below. Table 1: Classification of conditional forms ( Taylor 1997 ) Type Example Meaning and typical form Factual If prices go up, I sell my car. Content of the if -clause is presumed to be real and true. Form : present tense in the if -clause and present/future tense in the main clause Hypothetical If prices went up, I would sell my car . Content of the if -clause is regarded as a possibility. Form : past tense in the if -clause and a modal such as would (or similar form) plus a base verb form in the main clause. Counterfactual If prices had gone up, I would have sold my car . Content of the if -clause is regarded as not possible or contrary to the fact/current state of the world. Form : past perfect in the if -clause and a would (or another modal) combined with the perfect form of the verb in the main clause. Type Example Meaning and typical form Factual If prices go up, I sell my car. Content of the if -clause is presumed to be real and true. Form : present tense in the if -clause and present/future tense in the main clause Hypothetical If prices went up, I would sell my car . Content of the if -clause is regarded as a possibility. Form : past tense in the if -clause and a modal such as would (or similar form) plus a base verb form in the main clause. Counterfactual If prices had gone up, I would have sold my car . Content of the if -clause is regarded as not possible or contrary to the fact/current state of the world. Form : past perfect in the if -clause and a would (or another modal) combined with the perfect form of the verb in the main clause. Note : Examples come from Werth (1997 : 243–5). View Large Table 1: Classification of conditional forms ( Taylor 1997 ) Type Example Meaning and typical form Factual If prices go up, I sell my car. Content of the if -clause is presumed to be real and true. Form : present tense in the if -clause and present/future tense in the main clause Hypothetical If prices went up, I would sell my car . Content of the if -clause is regarded as a possibility. Form : past tense in the if -clause and a modal such as would (or similar form) plus a base verb form in the main clause. Counterfactual If prices had gone up, I would have sold my car . Content of the if -clause is regarded as not possible or contrary to the fact/current state of the world. Form : past perfect in the if -clause and a would (or another modal) combined with the perfect form of the verb in the main clause. Type Example Meaning and typical form Factual If prices go up, I sell my car. Content of the if -clause is presumed to be real and true. Form : present tense in the if -clause and present/future tense in the main clause Hypothetical If prices went up, I would sell my car . Content of the if -clause is regarded as a possibility. Form : past tense in the if -clause and a modal such as would (or similar form) plus a base verb form in the main clause. Counterfactual If prices had gone up, I would have sold my car . Content of the if -clause is regarded as not possible or contrary to the fact/current state of the world. Form : past perfect in the if -clause and a would (or another modal) combined with the perfect form of the verb in the main clause. Note : Examples come from Werth (1997 : 243–5). View Large In traditional instruction, learners are presented with three conditional types—factual, hypothetical, and counterfactual—and corresponding examples thereof. A number of factors contribute to the difficulty of conditionals for L2 learners ( Celce-Murcia and Larsen-Freeman 1999 ). First, because the conditional structure consists of two clauses, the syntactic aspects of subordination do not make them easy to acquire. Second, before being able to form conditionals correctly, learners need to master a number of other grammatical features, such as tense and aspect, modal auxiliaries, and negation. Lastly, the variety of types and contexts in which conditionals are used are often inadequately captured by typical ESL/EFL sources, as described in the following section. Representation of conditionals in EFL/ESL materials: Common problems In the majority of current ESL/EFL textbooks, the aspects of conditional form—as opposed to meaning—represent the prevalent instructional focus ( Jacobsen 2012 ). A typical classification (as referred to in Table 1 above) and accompanying examples do not explain how conditionals may be used in a variety of contexts, nor do they provide a clear link between different conditional types and their corresponding semantic motivations, or the exact linguistic environments where one type of conditional might be preferred over others ( Meunier 2008 ). (For a full overview of how conditionals are presented in textbooks, see Jacobsen 2012 ). Furthermore, a recent study based on corpus findings (Gabrielatos 2003, cited in Meunier 2008 ), that compared the traditional English language teaching terminology of conditional clauses against corpus evidence, found that 55 per cent of the authentic corpus examples did not fit into the core english language teaching (ELT) typology. Examples produced in naturally occurring contexts were found to be much more varied and multi-faceted than those captured by the traditional typology. For instance, using would forms in if -clauses is often treated as incorrect in textbooks; however, corpus data from different registers show that such usage is indeed permissible. In other words, traditional typologies appear to be at best schematic, and misleading and imprecise at worst, but still continue to be widely used in ESL/EFL materials. To counter these shortcomings, the CL approach has the tools to address conditional meaning and context usage, and make them fully observable and salient for the learner. CL approach to conditionals This study uses the CL analysis of English conditionals by Dancygier and Sweetser (2005) . Basing their approach on blending theory by Fauconnier and Turner (2002) , Dancygier and Sweetser claim that conditional if -clauses set up certain mental space structures. They analyze how such a set-up happens through manipulating various markers (e.g. if , when , unless ) and using different verb forms. Coherence and validity of conditional phrases depend on the successful configuration of all the elements constituting their structure, as individually construed by a given speaker. The authors’ approach takes into account the contextual needs and formal characteristics of conditionals as they fit in each context of use. Thus, Dancygier and Sweetser outline linguistic tools commonly employed by native English speakers to create felicitous conditional phrases, while also demonstrating the dependence of conditionals on the communicative context and the speaker’s stance toward a given situation. An important insight from Dancygier and Sweetser’s work is the idea of compositionality of conditionals: by analyzing combinations of regular and less regular conditional constructions and identifying patterns of inferential structure and metonymic reasoning involved, the authors essentially single out the elements constituting conditional constructions and make those aspects of meaning directly analyzable. Treating conditionals as meaning-form ‘packages’ ( Achard 2004 ) allows us to break down composite grammatical meanings into separate parts and subsequently highlight the form–meaning connections of those parts explicitly to learners. A careful adaptation of these CL insights into L2 materials has the ability to highlight individual aspects of conditional meaning and also outline patterns that learners could readily rely upon in the subsequent construction of their own conditional sentences. Explicit instruction and task-based language teaching CL theory can be incorporated into the classroom through explicit instruction, the effectiveness of which has been firmly established in prior research ( Norris and Ortega 2000 ; Ellis 2002 ; Spada and Tomita 2010 ). In particular, the underlying tenets of focus on form ( Long 2000 ), noticing ( Schmidt 2001 ), and consciousness-raising ( Sharwood Smith 1981 , 1993 , 2007 ) emphasize the effectiveness of helping the learner notice and understand key meaning motivations behind target forms. Focus on form or consciousness-raising is instrumental in helping learners notice linguistic cues, especially when targeting complex, polysemous, or potentially ambiguous linguistic forms (cf., Hulstijn & de Graaff 1994 ; Fordyce 2014 ; DeKeyser and Prieto Botana 2015 ). Furthermore, proponents of CL theory have suggested combining CL principles with selective focus on form ( Achard and Niemeier 2004 ; Cadierno 2008 ; Holme 2012 ). However, viable pedagogical solutions are needed to make these novel insights appear approachable to the learners. The paradigm of task-based (-supported 1 ) language teaching (TBLT) is highly compatible with the basic CL assumptions, the two providing a rich initial ground for further collaboration. TBLT has been gaining increased popularity over the course of the past two decades (e.g. Bygate et al. 2001 ; Robinson 2001 , 2005 ; Ellis 2003 ; Samuda and Bygate 2008 ), having emerged as a natural progression from communicative language teaching and shifting the pedagogic focus onto meaning-oriented and learner-centered instruction ( Van den Branden et al. 2009 ). While there exist a number of definitions for tasks, this study will rely on the comprehensive operationalization of task, provided in Ellis (2003 : 5–6): ‘A task is a workplan that requires learners to process language pragmatically in order to achieve an outcome… it requires them to give primary attention to meaning and to make use of their own linguistic resources, although the design of the task may predispose them to choose particular forms.’ Tasks provide learners with an opportunity to produce language in a context that resembles or aims to recreate an authentic language acquisition environment ( Révész 2009 : 438). While the focus on communication and meaning remains the primary element of TBLT, most researchers agree that focus on form incorporated into instruction in line with the learners’ developmental needs is crucial for boosting L2 acquisition ( Ellis 2003 ; Long 2000 ; Norris and Ortega 2000 ; Robinson 2001 ; Van den Branden et al. 2009 ). The value of using task-based methodology in combination with CL insights has been explicitly addressed in Cadierno (2008) . Furthermore, Cadierno and Robinson (2009) used pedagogic tasks to measure how the manipulation of cognitive complexity can facilitate the development of L2 construal patterns. Moder (2010) used targeted pedagogic tasks (based on prior corpus-based research) to teach usage patterns of the English like constructions. Finally, a number of experiments conducted at Georgetown University reported in Tyler (2012) used various versions of pedagogic tasks to deliver CL insights to learners. While it is a promising direction, the relative dearth of studies at this point in time does not allow us to make comprehensive conclusions about the best ways to combine TBLT with CL insights. This study thus aims to address this gap by complementing the applied CL focus with the use of pedagogic tasks. Task-supported instruction and consciousness-raising tasks The instructional context of this study is closer to ‘task-supported’ rather than ‘task-based’. In task-supported settings ( Ellis 2003 ), tasks are not the main medium of instruction but rather one of the key components of the general syllabus, providing learners with semi-authentic contexts for practicing target forms ( Samuda and Bygate 2008 : 60). As it will become clearer from the ‘Methods’ section, constraints of the graduate English for Academic Purposes (EAP) instructional context made it impossible to re-conceptualize the existing curriculum as entirely task-based. Relying on tasks as means of support, rather than the main component, of instruction was more feasible in the provided educational environment. In the present study, the task support was realized in form of the consciousness-raising tasks . Eckerth (2008a : 92) defines consciousness-raising tasks as ‘form-focused tasks’ that can be used as ‘a pedagogical device to direct learners’ attention to specific L2 forms while they are communicating in the L2’. The main purpose of consciousness-raising tasks is to give learners the opportunity to explore target features in a focused context while simultaneously attending to completing task guidelines. The efficacy of consciousness-raising tasks was proven in multiple L2 contexts ( Fotos and Ellis 1991 ; Fotos 1993 , 1994 ; De la Fuente 2006 ; Eckerth 2008a , 2008b ). Specifically, Fotos (1994) examined how grammar-focused consciousness-raising tasks affected proficiency gains in acquiring word order, finding that consciousness-raising tasks were a more powerful teaching method than formal instruction. In De la Fuente’s (2006) study on the efficacy of using consciousness-raising tasks for instruction of vocabulary, participants of the task-based group performed statistically significantly better on the delayed L2 vocabulary retrieval test than the traditional group participants, which suggests a positive long-term effect associated with consciousness-raising tasks. Eckerth (2008a ) reported that learners’ language scaffolding skills are necessary for successful completion of consciousness-raising tasks. Eckerth’s (2008b ) study focused on investigating the effects of dialogic tasks onto specific learning gains and on the process of task completion itself. The findings demonstrated that task completion yielded significant task-specific learning gains both immediately and after passage of some time. Furthermore, consciousness-raising tasks were associated with greater linguistic complexity in terms of both perception and production. These findings provide support for using consciousness-raising tasks as a way to deliver guided focus on form in L2 settings, that is, helping highlight CL insights in a meaningful usage context. METHODS Target structure The author conducted preliminary research in free online learner corpora (for full list with web addresses, see Appendix A in Jacobsen 2012 ) on most commonly occurring errors in conditionals, finding that the range of constructions carrying the predictive 2 function (e.g. If prices go up, I’ll sell/I sell my car/If prices went up, would sell my car/If prices had gone up, I would have sold my car )—as opposed to the pragmatic/speech act conditionals—represents the biggest challenge for L2 learners. In other words, learners experience the most difficulty (as exhibited through imprecise or erroneous tense sequences and/or other time indicators) when attempting to highlight hypothetical outcomes referring to the unreal situation in the past or the present. Accordingly, predictive conditionals were chosen as the target structure of this study. Research questions and hypotheses This study intended to examine the general effects of task-supported focus on form upon the instruction of conditionals, as well as the relative significance of CL insights included into the instructional treatment. The study addressed three research questions: Research Question 1 : Does task-supported instruction produce an effect for L2 development of conditional constructions? Research Question 2 : Does task-supported instruction with added CL insights produce an effect for L2 development of conditional constructions? Research Question 3 : Does the addition of CL insights into task-supported instruction produce greater L2 development of conditional constructions than task-supported instruction alone? The term ‘ CL insights ’ refers to CL explanations added to pedagogic materials. CL insights were part of instruction for only one of the groups (referred to as the cognitive group hereafter). The group that received instruction without the CL insights will be referred to as the task-supported group. Finally, effect for L2 development is operationalized as performance on the post-test and the delayed post-test (see section on study design). While there does exist considerable research pertaining to the effectiveness of CL in L2 teaching, its total scope and limitations do not yet allow for directional hypotheses; accordingly, it would be most appropriate to frame the research hypotheses as null statements. H1 : Task-supported instruction will not produce any effect for the L2 development of conditional constructions. H2 : Task-supported instruction with added CL insights will not produce an effect for L2 development of conditional constructions. H3 : The addition of CL insights into task-supported instruction will not produce greater L2 development of conditional constructions than task-supported instruction alone. Participants The data were collected in a graduate EAP program at a large university on the East Coast of the USA. The study included three main groups— cognitive , task-supported , and control . The subjects in the cognitive and task-supported groups were recruited from single intact graduate EAP classes, while the subjects in the control group came from two separate EAP classes. Most of the subjects were in their mid-20s. The total number of male participants across three groups was 36, and the total number of females was 21. The vast majority of participants came from Mainland China or Taiwan and spoke Mandarin Chinese as their first language, but there were a few participants with other L1s (see Table 2 ). A key limitation of this population sample is that the influence of L1 on the subjects’ prior conceptualization of the concept of conditionals was not assessed; such an intervention would require logistical arrangements incompatible with the curricular restrictions within the program. However, qualitative analysis of errors in learner corpora demonstrated that learners of all L1 backgrounds still struggle with forming grammatical tense sequences and assuming context-appropriate stance for expressing conditional meaning in English. To ensure that there were no differences in the prior level of mastery of English conditionals, a one-way analysis of variance (ANOVA) was administered on pretest data (see the ‘Results’ section for full report). Table 2: Study participants Cognitive group Task-supported group Control group Total number 17 18 22 Mean age 25 (SD = 5.6) 24 (SD = 4.7) 26 (SD = 4.2) Gender     Male 10 10 16     Female 7 8 6 L1     Mandarin Chinese 14 17 15     Arabic 1 1 4     Farsi 2 1     Portuguese (Brazil) 1     Urdu 1 Mean length of stay in the USA 3 months (SD = 2.3) 3 months (SD = 3.5) 7 months (SD = 2.9) Cognitive group Task-supported group Control group Total number 17 18 22 Mean age 25 (SD = 5.6) 24 (SD = 4.7) 26 (SD = 4.2) Gender     Male 10 10 16     Female 7 8 6 L1     Mandarin Chinese 14 17 15     Arabic 1 1 4     Farsi 2 1     Portuguese (Brazil) 1     Urdu 1 Mean length of stay in the USA 3 months (SD = 2.3) 3 months (SD = 3.5) 7 months (SD = 2.9) View Large Table 2: Study participants Cognitive group Task-supported group Control group Total number 17 18 22 Mean age 25 (SD = 5.6) 24 (SD = 4.7) 26 (SD = 4.2) Gender     Male 10 10 16     Female 7 8 6 L1     Mandarin Chinese 14 17 15     Arabic 1 1 4     Farsi 2 1     Portuguese (Brazil) 1     Urdu 1 Mean length of stay in the USA 3 months (SD = 2.3) 3 months (SD = 3.5) 7 months (SD = 2.9) Cognitive group Task-supported group Control group Total number 17 18 22 Mean age 25 (SD = 5.6) 24 (SD = 4.7) 26 (SD = 4.2) Gender     Male 10 10 16     Female 7 8 6 L1     Mandarin Chinese 14 17 15     Arabic 1 1 4     Farsi 2 1     Portuguese (Brazil) 1     Urdu 1 Mean length of stay in the USA 3 months (SD = 2.3) 3 months (SD = 3.5) 7 months (SD = 2.9) View Large The vast majority of the participants first came to the USA a few months prior to the beginning of data collection; for all students, enrolling in this graduate program was their first educational experience in the English-speaking world. All subjects previously studied English with various degrees of instructional intensity in secondary school and possibly college; however, due to lack of uniform reporting on the subjects’ prior educational contexts, their proficiency was captured only through standardized test scores: everyone earned a minimum of 80 on the TOEFL iBT prior to being admitted to university. Table 2 represents basic biographic information related to the study participants. The treatment types received by each group were as follows: the control group ( N = 22) received no explicit classroom instruction targeting conditionals and only completed the three tests; the task-supported group ( N = 18) received instruction targeting conditionals using the traditional teacher-facilitated explanation and pedagogic tasks, that is, materials for task-supported group did not include explanations guided by CL insights; the cognitive group ( N = 17) received teacher-facilitated presentations of conditionals as well as pedagogic tasks. Teacher-used instructional materials did include CL explanations of the target form. Instructional procedure The entire process of data collection took six weeks. During the first week, all three groups completed pretests. The pretest scores were approximately the same across all three groups (25, 24.5, and 23.5 for the cognitive, task-supported, and control groups, respectively; also see the section on test administration and scoring for details on one-way ANOVA). The treatment for the cognitive and task-supported groups took place over the course of three weeks. Following its completion, immediate and delayed (10 days after the end of instruction) post-tests were administered to measure relative gains that may have happened in the course of the treatment. Each of the tests took approximately 50 minutes to complete. To collect qualitative data in the first week following the completion of treatment, three and four subjects from task-supported and cognitive groups, respectively, were interviewed using the method of retrospective interview . The details and findings pertaining to the retrospective interviews were addressed in full in Jacobsen (2015) and will therefore be omitted here. The full study design is available in Appendix A. Tests The tests created for this study attempted to address both implicit and explicit knowledge and reflected production and comprehension aspects of conditional usage. They consisted of four distinct parts: controlled production, free production, ‘grammaticality judgment/comprehension: pictures’, and ‘grammaticality judgment/comprehension: sentences’. Controlled and free production parts, where students were asked to fill in the blanks and describe a picture respectively, thus demonstrating how they used the target form in specific contexts, aimed to target primarily implicit knowledge ( Norris and Ortega 2000 ; Ellis 2009 ). Grammaticality judgment/comprehension parts, on the other hand, were supposed to target explicit as well as implicit knowledge (cf., Fotos 1994 ; Han and Ellis 1998 ; Ellis 2004 ). Specifically, Ellis (2004 : 249) referred to grammaticality judgment tests as a ‘favored method of investigating L2 explicit knowledge’. The maximum number of points on each test was 49, as demonstrated in the cross-sectional breakdown below: controlled production (13 items, 29 points); free production (4 items, 8 points); grammaticality judgment and comprehension: pictures (4 items, 4 points); grammaticality judgment and comprehension: sentences (8 items, 8 points). Test forms included discourse excerpts representing different genres and registers from the Corpus of American Contemporary English ( http://corpus.byu.edu/coca /) to recreate authentic usage patterns of conditionals. Test images were obtained through searches on Web sites with Common Creative Licenses. Comprehensive answer keys were created for each version of pre-, post-, and delayed post-test. The answer keys reflected the opinions of native speakers obtained during the piloting process. Since most test items were taken from authentic discourse sources, the opinions of native speakers were compared with the original verb tenses from the authentic corpus excerpts; if the sources disagreed, the item was discarded. All test items were then piloted with native and non-native speakers. Only the non-ambiguous items (those not causing difficulty for either native or non-native speakers) were used for the final test forms. Each of the four test subsections had three possible versions (A, B, and C), and those were counterbalanced for each test installment (pretest, post-test, and delayed post-test). Instructional materials Each treatment group received three teacher-facilitated PowerPoint presentations (one per each week of instruction) and six pedagogic tasks (two per each week of instruction). The length of each of the presentations described below was about 50 minutes, or the total of 150 minutes divided across six sessions; in other words, both groups received explicit pedagogical treatment of the same length. However, even though both groups addressed metalinguistic aspects of language, they did so in different ways: that is, each group’s presentation content was condition specific (see next two sections). The pedagogic tasks (see separate section) were the same for both treatment groups. Cognitive group All three cognitive presentations were informed by a single CL account of conditionals ( Dancygier and Sweetser 2005 ) and aimed to represent the CL insights in a learner-accessible manner. The first presentation focused on exposing students to the meaning-centered and compositional reality of language. The corresponding mental steps needed for creating conditionals were presented to students as follows: Start with thinking about the background knowledge in question. What is realistic? What is the possible or projected time of condition ( if -clause)? What is the possible or projected time of result (main clause)? Pick the tenses that will reflect such an arrangement. Check if your sentence makes sense in the given context. (The full version of the first PowerPoint (PPT) is available as appendix in Jacobsen 2015 .) The cognitive chart (available in Appendix B) was designed to serve as a one-stop reminder of all these steps, mirroring the cognitive processes necessary for creating felicitous conditionals. Cognitive group participants could use it during the teacher-facilitated instruction and during the pedagogic tasks that they completed in class. Even though students were encouraged to bring their copies of the cognitive chart to every class, it was noted that not all participants made active use of it. While this supplementary resource was available to everyone, no conclusions can be drawn about its distinct use and/or effectiveness in the context of this study. Furthermore, the second presentation, available in Appendix C, demonstrated the underlying meanings of English tenses and how they are used to express a variety of semantic configurations. The third presentation aimed at showing the subjects how conditionals are shaped within specific usage contexts and how surrounding context-specific information might affect the tense composition of either of the two clauses. The cognitive PPT presentations thus explained the general system behind the notion of conditional sentences by focusing on meaning behind every single grammatical form and its usage contexts. In short, the students had to consider and keep in mind the following characteristics when pondering the compositional meaning and implications behind a given conditional sentence: Tenses in surrounding sentences. Perspective of speaker who is narrating: is it the same person as a given character? Is s/he still within the zone of control or already outside it? Time markers such as adverbs, adjectives, providing an idea of what time is referenced. Appropriateness of forms for corresponding discourse context. Task-supported group The task-supported group received a traditional presentation of the English conditionals informed by L2 textbooks and ‘ESL Grammar Book’ ( Celce-Murcia and Larsen-Freeman 1999 ). During the first week of instructional treatment, learners were given an overview of all three conditional types with subtypes as informed by Celce-Murcia and Larsen-Freeman’s extended classification. During the second week, the focus was on tense combinations in all types of conditional sentences. The tense sequences were highlighted according to the explanations and descriptions provided in the traditional ESL materials. During the third week, the focus was on practicing using these forms in various contexts, as provided by the textbooks. While they received a traditionally balanced presentation of various aspects of the conditional form, subjects of the task-supported group were not asked to engage in conscious reflections on meaning behind different tense pairings beyond the range discussed in textbooks and beyond the meaning context provided by pedagogic tasks. Pedagogic tasks Both treatment groups completed a shared set of pedagogic tasks to support the teacher-facilitated PPT presentations. The conditions of these tasks aimed to resemble the authentic context of target language use as much as possible. All tasks were consciousness-raising in the sense that they directed subjects’ attention toward the use of conditionals, and conditionals were crucial for their successful completion. All tasks and their key characteristics are captured in Table 3 . The results of this instructional intervention are reported in the next section. RESULTS Descriptive statistics One-way ANOVA administered on pretest data ensured that all groups were at the same proficiency level (cognitive vs. task-supported: p = .904, standard error (SE) = 1.6; task-supported vs. control: p = .86, SE = 1.5; cognitive vs. control: p = .60, SE= 1.55). Descriptive statistics for all three groups by testing task and an overview of mean overall scores obtained across three tests by all groups are presented in Tables 4 and 5 , respectively, below. Table 3: Pedagogic tasks used in the study Time Task order and name Knowledge/skills addressed through task Foci Gist of task WEEK 1 1 Sentence strips Sequence of tenses between clauses, time markers Controlled production, comprehension; grammar task In groups of 2–3, students discuss meaning of sentence strips representing two parts of conditionals (with blanks for verb tenses). Then students match if -clauses with corresponding main clauses, supplying the correct verb tenses in the blanks and providing an appropriate rationale for their choice. 2 Understanding background knowledge Speaker knowledge Consciousness-raising, comprehension; grammar task Students are given four conditional sentences and asked to explain background knowledge that had to precede formation of each conditional, tying the rationale to form (tense choices). WEEK 2 3 Hedging in data commentaries Tenses and context knowledge Controlled production, comprehension; grammar task Adapted from Swales and Feak (2004) , this task elicited students’ written data commentary of available statistical information within provided table. Students had to frame their data commentary as hypotheses and predictions using conditionals. 4 Global warming causes Tenses; speaker and context knowledge Comprehension, controlled production; grammar task Students discuss causes and effects of global warming using conditionals. For content knowledge, students relied on materials and class discussions from earlier in the semester. WEEK 3 5 Text repair: break-up letter Tenses, context knowledge, time markers Consciousness-raising, controlled production; communicative task Students identify ill-formed conditionals in a break-up letter and rewrite the letter in own way supplying correct forms based on understanding/interpretation of local semantic context. 6 Seating chart task Tenses, speaker knowledge Free production; grammar task Working in groups, students try to seat people of different and sometimes conflicting backgrounds around one dining table using conditionals. The conditions listed could not be satisfied in their entirety, so students had to try to satisfy as many conditions as they possibly could. Time Task order and name Knowledge/skills addressed through task Foci Gist of task WEEK 1 1 Sentence strips Sequence of tenses between clauses, time markers Controlled production, comprehension; grammar task In groups of 2–3, students discuss meaning of sentence strips representing two parts of conditionals (with blanks for verb tenses). Then students match if -clauses with corresponding main clauses, supplying the correct verb tenses in the blanks and providing an appropriate rationale for their choice. 2 Understanding background knowledge Speaker knowledge Consciousness-raising, comprehension; grammar task Students are given four conditional sentences and asked to explain background knowledge that had to precede formation of each conditional, tying the rationale to form (tense choices). WEEK 2 3 Hedging in data commentaries Tenses and context knowledge Controlled production, comprehension; grammar task Adapted from Swales and Feak (2004) , this task elicited students’ written data commentary of available statistical information within provided table. Students had to frame their data commentary as hypotheses and predictions using conditionals. 4 Global warming causes Tenses; speaker and context knowledge Comprehension, controlled production; grammar task Students discuss causes and effects of global warming using conditionals. For content knowledge, students relied on materials and class discussions from earlier in the semester. WEEK 3 5 Text repair: break-up letter Tenses, context knowledge, time markers Consciousness-raising, controlled production; communicative task Students identify ill-formed conditionals in a break-up letter and rewrite the letter in own way supplying correct forms based on understanding/interpretation of local semantic context. 6 Seating chart task Tenses, speaker knowledge Free production; grammar task Working in groups, students try to seat people of different and sometimes conflicting backgrounds around one dining table using conditionals. The conditions listed could not be satisfied in their entirety, so students had to try to satisfy as many conditions as they possibly could. View Large Table 3: Pedagogic tasks used in the study Time Task order and name Knowledge/skills addressed through task Foci Gist of task WEEK 1 1 Sentence strips Sequence of tenses between clauses, time markers Controlled production, comprehension; grammar task In groups of 2–3, students discuss meaning of sentence strips representing two parts of conditionals (with blanks for verb tenses). Then students match if -clauses with corresponding main clauses, supplying the correct verb tenses in the blanks and providing an appropriate rationale for their choice. 2 Understanding background knowledge Speaker knowledge Consciousness-raising, comprehension; grammar task Students are given four conditional sentences and asked to explain background knowledge that had to precede formation of each conditional, tying the rationale to form (tense choices). WEEK 2 3 Hedging in data commentaries Tenses and context knowledge Controlled production, comprehension; grammar task Adapted from Swales and Feak (2004) , this task elicited students’ written data commentary of available statistical information within provided table. Students had to frame their data commentary as hypotheses and predictions using conditionals. 4 Global warming causes Tenses; speaker and context knowledge Comprehension, controlled production; grammar task Students discuss causes and effects of global warming using conditionals. For content knowledge, students relied on materials and class discussions from earlier in the semester. WEEK 3 5 Text repair: break-up letter Tenses, context knowledge, time markers Consciousness-raising, controlled production; communicative task Students identify ill-formed conditionals in a break-up letter and rewrite the letter in own way supplying correct forms based on understanding/interpretation of local semantic context. 6 Seating chart task Tenses, speaker knowledge Free production; grammar task Working in groups, students try to seat people of different and sometimes conflicting backgrounds around one dining table using conditionals. The conditions listed could not be satisfied in their entirety, so students had to try to satisfy as many conditions as they possibly could. Time Task order and name Knowledge/skills addressed through task Foci Gist of task WEEK 1 1 Sentence strips Sequence of tenses between clauses, time markers Controlled production, comprehension; grammar task In groups of 2–3, students discuss meaning of sentence strips representing two parts of conditionals (with blanks for verb tenses). Then students match if -clauses with corresponding main clauses, supplying the correct verb tenses in the blanks and providing an appropriate rationale for their choice. 2 Understanding background knowledge Speaker knowledge Consciousness-raising, comprehension; grammar task Students are given four conditional sentences and asked to explain background knowledge that had to precede formation of each conditional, tying the rationale to form (tense choices). WEEK 2 3 Hedging in data commentaries Tenses and context knowledge Controlled production, comprehension; grammar task Adapted from Swales and Feak (2004) , this task elicited students’ written data commentary of available statistical information within provided table. Students had to frame their data commentary as hypotheses and predictions using conditionals. 4 Global warming causes Tenses; speaker and context knowledge Comprehension, controlled production; grammar task Students discuss causes and effects of global warming using conditionals. For content knowledge, students relied on materials and class discussions from earlier in the semester. WEEK 3 5 Text repair: break-up letter Tenses, context knowledge, time markers Consciousness-raising, controlled production; communicative task Students identify ill-formed conditionals in a break-up letter and rewrite the letter in own way supplying correct forms based on understanding/interpretation of local semantic context. 6 Seating chart task Tenses, speaker knowledge Free production; grammar task Working in groups, students try to seat people of different and sometimes conflicting backgrounds around one dining table using conditionals. The conditions listed could not be satisfied in their entirety, so students had to try to satisfy as many conditions as they possibly could. View Large Table 4: Descriptive statistics by testing task Pretest means Post-test means Delayed post-test means Testing task Cog Task Control Cog Task Control Cog Task Control Controlled production 13.59 12.61 13.18 22.06 18.11 14.82 21.29 16.56 14.32 Free production 3.82 4.11 3.55 6.24 5.56 3.77 6.29 5.61 4.45 Comprehension: pictures 2.65 2.83 2.91 3.24 2.94 2.77 2.71 2.50 2.50 Comprehension: sentences 5.12 4.89 3.95 6.65 4.94 4.23 6.24 5.11 4.14 Pretest means Post-test means Delayed post-test means Testing task Cog Task Control Cog Task Control Cog Task Control Controlled production 13.59 12.61 13.18 22.06 18.11 14.82 21.29 16.56 14.32 Free production 3.82 4.11 3.55 6.24 5.56 3.77 6.29 5.61 4.45 Comprehension: pictures 2.65 2.83 2.91 3.24 2.94 2.77 2.71 2.50 2.50 Comprehension: sentences 5.12 4.89 3.95 6.65 4.94 4.23 6.24 5.11 4.14 View Large Table 4: Descriptive statistics by testing task Pretest means Post-test means Delayed post-test means Testing task Cog Task Control Cog Task Control Cog Task Control Controlled production 13.59 12.61 13.18 22.06 18.11 14.82 21.29 16.56 14.32 Free production 3.82 4.11 3.55 6.24 5.56 3.77 6.29 5.61 4.45 Comprehension: pictures 2.65 2.83 2.91 3.24 2.94 2.77 2.71 2.50 2.50 Comprehension: sentences 5.12 4.89 3.95 6.65 4.94 4.23 6.24 5.11 4.14 Pretest means Post-test means Delayed post-test means Testing task Cog Task Control Cog Task Control Cog Task Control Controlled production 13.59 12.61 13.18 22.06 18.11 14.82 21.29 16.56 14.32 Free production 3.82 4.11 3.55 6.24 5.56 3.77 6.29 5.61 4.45 Comprehension: pictures 2.65 2.83 2.91 3.24 2.94 2.77 2.71 2.50 2.50 Comprehension: sentences 5.12 4.89 3.95 6.65 4.94 4.23 6.24 5.11 4.14 View Large Table 5: Descriptive statistics of mean test scores across three groups Score Group Mean SD N Pretest score Cognitive 25.18 5.27 17 Task 24.44 4.33 18 Control 23.59 4.83 22 Post-test score Cognitive 38.06 3.25 17 Task 31.67 5.35 18 Control 25.59 4.85 22 Delayed post-test score Cognitive 36.53 3 17 Task 29.78 4.44 18 Control 25.45 5.71 22 Score Group Mean SD N Pretest score Cognitive 25.18 5.27 17 Task 24.44 4.33 18 Control 23.59 4.83 22 Post-test score Cognitive 38.06 3.25 17 Task 31.67 5.35 18 Control 25.59 4.85 22 Delayed post-test score Cognitive 36.53 3 17 Task 29.78 4.44 18 Control 25.45 5.71 22 View Large Table 5: Descriptive statistics of mean test scores across three groups Score Group Mean SD N Pretest score Cognitive 25.18 5.27 17 Task 24.44 4.33 18 Control 23.59 4.83 22 Post-test score Cognitive 38.06 3.25 17 Task 31.67 5.35 18 Control 25.59 4.85 22 Delayed post-test score Cognitive 36.53 3 17 Task 29.78 4.44 18 Control 25.45 5.71 22 Score Group Mean SD N Pretest score Cognitive 25.18 5.27 17 Task 24.44 4.33 18 Control 23.59 4.83 22 Post-test score Cognitive 38.06 3.25 17 Task 31.67 5.35 18 Control 25.59 4.85 22 Delayed post-test score Cognitive 36.53 3 17 Task 29.78 4.44 18 Control 25.45 5.71 22 View Large Table 5 shows that mean pretest scores were quite similar for all three groups, with 25.18, 24.44, and 23.59 for cognitive, task-supported, and control group, respectively, the average standard deviation (SD) among all three groups being 4.778. On the post-test, average score of the cognitive group participants was 38.06, the one for the task-supported group participants was 31.67, and the one for the control group participants was 25.59, the SD for the entire sample being 6.871. Thus, the average score of cognitive group participants expanded by 13 points, that of task-supported group participants grew by 7 points, while that of control group increased by 2 points. The delayed post-test scores were 36.53, 29.78, and 25.45 for the cognitive, task-supported, and control groups, respectively, which demonstrate an average loss of about 2 points for both treatment groups; the score of control group stayed approximately the same, declining only by an average of one tenth of a point between the post-test and the delayed post-test. ANOVA comparison for all three groups In the present study, a repeated-measures within-subjects ANOVA was chosen for the main statistical procedure because the same subjects contributed to test means at different points of the experiment. Mauchly’s test of sphericity was conducted to ensure that the variances of the differences between conditions were equal. The p -value obtained on this test (.448) was not significant; therefore, the condition of sphericity could be considered satisfied for the present data and study context. Normality assumptions (skewness and kurtosis) were also met for this data set. The data in Table 6 below provide an overview of tests of within-subjects effects, while Table 7 reports the results of multivariate tests. Table 6: Repeated measures ANOVA: Tests of within-subjects effects Source df Mean square F Sig. ** Time Sphericity Assumed 2 881.486 70.66 .000 Greenhouse-Geisser 1.942 907.824 70.66 .000 Huynh-Feldt 2.000 881.486 70.66 .000 Lower-bound 1.000 1762.973 70.66 .000 Time * GROUP Sphericity Assumed 4 170.188 13.64 .000 Greenhouse-Geisser 3.884 175.273 13.64 .000 Huynh-Feldt 4.000 170.188 13.64 .000 Lower-bound 2.000 340.376 13.64 .000 Source df Mean square F Sig. ** Time Sphericity Assumed 2 881.486 70.66 .000 Greenhouse-Geisser 1.942 907.824 70.66 .000 Huynh-Feldt 2.000 881.486 70.66 .000 Lower-bound 1.000 1762.973 70.66 .000 Time * GROUP Sphericity Assumed 4 170.188 13.64 .000 Greenhouse-Geisser 3.884 175.273 13.64 .000 Huynh-Feldt 4.000 170.188 13.64 .000 Lower-bound 2.000 340.376 13.64 .000 ** p < .01. View Large Table 6: Repeated measures ANOVA: Tests of within-subjects effects Source df Mean square F Sig. ** Time Sphericity Assumed 2 881.486 70.66 .000 Greenhouse-Geisser 1.942 907.824 70.66 .000 Huynh-Feldt 2.000 881.486 70.66 .000 Lower-bound 1.000 1762.973 70.66 .000 Time * GROUP Sphericity Assumed 4 170.188 13.64 .000 Greenhouse-Geisser 3.884 175.273 13.64 .000 Huynh-Feldt 4.000 170.188 13.64 .000 Lower-bound 2.000 340.376 13.64 .000 Source df Mean square F Sig. ** Time Sphericity Assumed 2 881.486 70.66 .000 Greenhouse-Geisser 1.942 907.824 70.66 .000 Huynh-Feldt 2.000 881.486 70.66 .000 Lower-bound 1.000 1762.973 70.66 .000 Time * GROUP Sphericity Assumed 4 170.188 13.64 .000 Greenhouse-Geisser 3.884 175.273 13.64 .000 Huynh-Feldt 4.000 170.188 13.64 .000 Lower-bound 2.000 340.376 13.64 .000 ** p < .01. View Large Table 7: Repeated-measures ANOVA: Multivariate tests for performance across time in three groups Effect Value F Sig. Partial Eta Squared Time Pillai's Trace 0.74 77.08 .000 0.74 Wilks' Lambda 0.26 77.08 .000 0.74 Hotelling's Trace 2.91 77.08 .000 0.74 Roy's Largest Root 2.91 77.08 .000 0.74 Time * GROUP Pillai's Trace 0.54 9.86 .000 0.27 Wilks' Lambda 0.47 12.12 .000 0.31 Hotelling's Trace 1.11 14.44 .000 0.36 Roy's Largest Root 1.10 29.7 .000 0.52 Effect Value F Sig. Partial Eta Squared Time Pillai's Trace 0.74 77.08 .000 0.74 Wilks' Lambda 0.26 77.08 .000 0.74 Hotelling's Trace 2.91 77.08 .000 0.74 Roy's Largest Root 2.91 77.08 .000 0.74 Time * GROUP Pillai's Trace 0.54 9.86 .000 0.27 Wilks' Lambda 0.47 12.12 .000 0.31 Hotelling's Trace 1.11 14.44 .000 0.36 Roy's Largest Root 1.10 29.7 .000 0.52 View Large Table 7: Repeated-measures ANOVA: Multivariate tests for performance across time in three groups Effect Value F Sig. Partial Eta Squared Time Pillai's Trace 0.74 77.08 .000 0.74 Wilks' Lambda 0.26 77.08 .000 0.74 Hotelling's Trace 2.91 77.08 .000 0.74 Roy's Largest Root 2.91 77.08 .000 0.74 Time * GROUP Pillai's Trace 0.54 9.86 .000 0.27 Wilks' Lambda 0.47 12.12 .000 0.31 Hotelling's Trace 1.11 14.44 .000 0.36 Roy's Largest Root 1.10 29.7 .000 0.52 Effect Value F Sig. Partial Eta Squared Time Pillai's Trace 0.74 77.08 .000 0.74 Wilks' Lambda 0.26 77.08 .000 0.74 Hotelling's Trace 2.91 77.08 .000 0.74 Roy's Largest Root 2.91 77.08 .000 0.74 Time * GROUP Pillai's Trace 0.54 9.86 .000 0.27 Wilks' Lambda 0.47 12.12 .000 0.31 Hotelling's Trace 1.11 14.44 .000 0.36 Roy's Largest Root 1.10 29.7 .000 0.52 View Large Table 6 shows that the p value is <.001, demonstrating that the differences between the subjects’ mean scores on all three tests were statistically significant for each group pairing (i.e. between cognitive and task-supported groups, between task-supported and control groups, and between cognitive and control groups). As can be seen from Table 7 , multivariate tests focusing on the interaction between time and group variables demonstrate medium-strength effect sizes (partial eta squared). To check the specific areas of interaction between groups, Scheffe’s post hoc comparison was conducted; the results are represented in Table 8 below. Table 8: Repeated measures ANOVA: Multiple comparisons, Scheffe’s post hoc test (I) Group (J) Group Mean difference (I-J) Standard Error Sig. Cognitive Task 4.63 * 1.251 .002 Control 8.38 * 1.195 .000 Task Cognitive −4.63 * 1.251 .002 Control 3.75 * 1.176 .009 Control Cognitive −8.38 * 1.195 .000 Task −3.75 * 1.176 .009 (I) Group (J) Group Mean difference (I-J) Standard Error Sig. Cognitive Task 4.63 * 1.251 .002 Control 8.38 * 1.195 .000 Task Cognitive −4.63 * 1.251 .002 Control 3.75 * 1.176 .009 Control Cognitive −8.38 * 1.195 .000 Task −3.75 * 1.176 .009 * p < .01. View Large Table 8: Repeated measures ANOVA: Multiple comparisons, Scheffe’s post hoc test (I) Group (J) Group Mean difference (I-J) Standard Error Sig. Cognitive Task 4.63 * 1.251 .002 Control 8.38 * 1.195 .000 Task Cognitive −4.63 * 1.251 .002 Control 3.75 * 1.176 .009 Control Cognitive −8.38 * 1.195 .000 Task −3.75 * 1.176 .009 (I) Group (J) Group Mean difference (I-J) Standard Error Sig. Cognitive Task 4.63 * 1.251 .002 Control 8.38 * 1.195 .000 Task Cognitive −4.63 * 1.251 .002 Control 3.75 * 1.176 .009 Control Cognitive −8.38 * 1.195 .000 Task −3.75 * 1.176 .009 * p < .01. View Large Results presented in Table 8 suggest that the relationships between all group pairings were statistically significant with consistently observable p < .01. A visual representation of the how participants of each group scored across all three tests is given in Figure 1 below. Figure 1: View largeDownload slide Means of test performance over time Figure 1: View largeDownload slide Means of test performance over time Furthermore, ANOVAs to compare performance on each testing task were produced to determine the areas of greatest and least improvement over the period of the study. ANOVA for controlled production. Overall differences for controlled production were found to be statistically significant (between-subjects effects for interaction between time and group: F = 8.4, p = .001, partial eta-squared = 0.237). To avoid possible redundancy in discussion, only the data demonstrating actual interactions between groups will be reported. Scheffe’s post hoc test showed that the differences between the cognitive and task-supported groups and the cognitive and control groups were statistically significant, while the differences between the task-supported and the control groups were not. ANOVA for free production. For the free production part of the tests (between-subjects effects for interaction between time and group: F = 3.5, p = .01, partial eta squared = 0.11), the differences between the cognitive and the control groups, as well as between the task-supported and the control groups, were found to be statistically significant, while the difference between the cognitive and the task-supported groups was not. ANOVA for ‘comprehension: pictures’. For the ‘comprehension: pictures’ part, none of the interactions between groups proved to be statistically significant (between-subjects effects for interaction between time and group: F = 1.0, p = .4, partial eta squared = 0.03). The lack of significance may be related to the small number of items in this part of the test (four total). ANOVA for ‘comprehension: sentences’. For the final part of the test, ‘comprehension: sentences’, the differences between all group pairings proved to be statistically significant, albeit with marginal effect size ( F = 3.4, p = .04, partial eta squared = 0.06). Accordingly, these findings suggest that production parts of the tests—in particular, the controlled production—were the areas of biggest improvement over the course of the study. t -test comparison between cognitive and task-supported groups on overall test gain scores The descriptive statistics pertaining to the overall test scores and the production scores indicate clear differences between post-test outcomes in cognitive as opposed to task-supported groups. Descriptive statistics for the overall gains scores between the cognitive and task-supported groups are presented in Table 9 below. Table 9: Descriptive statistics for overall gains scores: Cognitive vs. task-supported groups Group N Mean SD SE mean Cognitive 17 12.82 5.01 1.21 Task-supported 18 7.28 5.43 1.28 Group N Mean SD SE mean Cognitive 17 12.82 5.01 1.21 Task-supported 18 7.28 5.43 1.28 View Large Table 9: Descriptive statistics for overall gains scores: Cognitive vs. task-supported groups Group N Mean SD SE mean Cognitive 17 12.82 5.01 1.21 Task-supported 18 7.28 5.43 1.28 Group N Mean SD SE mean Cognitive 17 12.82 5.01 1.21 Task-supported 18 7.28 5.43 1.28 View Large As can be seen from the table, the mean overall gain score for the cognitive group was 12.82, while the mean overall gain score for the task-supported group was 7.28, with the SD being in the low 5+ range (SE = 1.216 and 1.280 for the cognitive and the task-supported groups, respectively). The overview of the t -test comparison between these trends is provided in Table 10 . Table 10: T-test on overall gains scores: Cognitive and task-supported groups t df Sig. (2-tailed) Mean difference 3.13 33 .004 5.55 t df Sig. (2-tailed) Mean difference 3.13 33 .004 5.55 View Large Table 10: T-test on overall gains scores: Cognitive and task-supported groups t df Sig. (2-tailed) Mean difference 3.13 33 .004 5.55 t df Sig. (2-tailed) Mean difference 3.13 33 .004 5.55 View Large An independent samples t -test procedure was chosen because differences between two distinct groups were being measured through comparing individual/independent scores on two sets of tests. Levene’s test for equality of variances ( p =.826; F = 0.049) confirmed that equal variances can be assumed for this analysis. The two-tailed test for equality of means proved to be statistically significant with p = .004 with t (33) = 3.133. Summary of results The cognitive group obtained greater production score and overall test score gains between the pretest and the post-test, than both the task-supported and the control groups did. Participants of the cognitive and the task-supported groups were able to obtain greater test score gains between the pretest and the post-test, than the participants of the control group, thus providing support for the value of task-based classroom instruction for the acquisition of English conditionals. DISCUSSION Meaning of results This section will address answers to the research questions and general implications of the findings of this study. Research Question 1 : Does task-supported instruction produce an effect for L2 development of conditional constructions? Research Question 2 : Does task-supported instruction with added CL insights produce an effect for L2 development of conditional constructions? Research Question 3 : Does the addition of CL insights into task-supported instruction produce greater L2 development of conditional constructions than task-supported instruction alone? In response to the first research question, the initially assumed null hypothesis can be rejected, as the task-supported group outperformed the control group on both the post-test and the delayed post-test: the differences between the scores of these two groups were statistically significant. The second null hypothesis (research question 2) can also be rejected: statistical analyses demonstrated that the cognitive group outperformed the control group on both the post-test and the delayed post-test. In sum, both treatment groups outperformed the control group, thus demonstrating that explicit instruction of conditionals supported by pedagogic tasks is more effective for L2 development of this target form than no instruction. Furthermore, in response to the third question, the statistical analyses demonstrated that the cognitive group outperformed the task-supported group on both the post-test and the delayed post-test. In addition to general trends regarding effectiveness of instruction types, the analysis of post-test data revealed the following trends worth highlighting and discussing in greater detail: Changes in comprehension scores . Prior research has generally assumed that comprehension knowledge is supposed to precede production and, accordingly, the receptive knowledge generally exceeds the productive knowledge for L2 learners ( Lightbown 1985 , 2000 ; Ellis 2004 ). Judging by pretest scores on comprehension sections, all groups initially were at approximately the same level. After the instructional treatment, almost all of the cognitive group participants made consistent, albeit small (around 1–2 points), gains on the comprehension sections, while for participants of the task-supported and control groups, the gains in comprehension scores were not as consistent. Even though the numbers of comprehension items in the tests were too small for inferential statistics, this small-scale trend is suggestive of improved comprehension of conditionals that resulted from exposure to the CL treatment. Changes in production scores . Production parts of the tests were where most improvement took place between the pretest and the post-test, suggesting that the instructional treatment was instrumental for promoting production. It can be inferred that pedagogical tasks and the teacher-facilitated instructional PowerPoint presentations were effective methods for improving the participants’ production skills. Between the two treatment groups, the cognitive group outperformed the task-supported group on the production parts of the test, and the difference was calculated to be statistically significant with p < .01. The implication of this finding is that the combination of cognitive and task-supported treatment was more effective in improving the participants’ production of conditionals than the task-supported treatment alone. Conducting further research focusing specifically on production aspects of conditional phrases would be necessary, before any further, more certain conclusions could be drawn in this regard. Changes in overall test gain scores . The ANOVA and accompanying comparisons measuring gains between tests demonstrated that the difference between the gains made by participants of the two treatment groups was statistically significant with p < .001. This statistical finding suggests that the combination of the CL presentation of conditionals complemented by pedagogical tasks was generally more effective than the task-supported treatment alone: the gains made by the cognitive group participants were almost double compared to those of the task-supported group participants. It would be fair to hypothesize that the presentation of conditionals to the cognitive group was more cohesive and meaningful because it had a consistent foundation in the shape of CL presentation of language. Accordingly, the performance of the cognitive group may be attributed to the presence of the general focus on meaning and the CL insights emphasizing the role of local context and speaker’s background knowledge when creating a conditional phrase. The benefits associated with focus on meaning were also highlighted by the CL group subjects participating in retrospective interviews (for a full discussion of qualitative data pertaining to this study, see Jacobsen (2015) . Furthermore, since both treatment groups demonstrated improvement on test scores between the pretest and the post-test and since tasks were the instructional element shared by both groups, it is highly likely that tasks were the crucial instructional aspect contributing to L2 development within both treatment groups in comparison with the control group. Summative findings from the retrospective interview data demonstrated that students from both groups explicitly favored tasks that were relatively more complex (i.e. some type of cognitive effort was necessary for their completion), included relatively richer contextual knowledge, and required production and group work. While this study did not explicitly test for effects of task complexity on L2 development, this finding appears to be consistent with Robinson’s (2005 , 2011 ) research on cognitive task demands and their effects upon L2 development. Taking these trends into a broader theoretical context, it is important to highlight the fact that focus on form alone (as received by the task-supported group) was not demonstrated to be as effective as the focus on form combined with CL insights. While the support for the instruction (consciousness-raising tasks) mattered and produced gains for both treatment groups, the underlying theory of language assumed in teacher-facilitated presentation played a further facilitative role for L2 development. It is possible that tasks helped further contextualize the meaning conveyed through the cognitive explanation, and thus the use of tasks may have further enhanced the performance of the cognitive group. However, since we did not have a cognitive-only (no tasks) group in this study, it is not possible to specify the precise nature of interaction between the two variables. (Findings discussed in Jacobsen (2015) also supported the use of pedagogic tasks for cognitive-informed and traditional instruction.) Since the cognitive group outperformed the task-supported group, it can be implied that the CL theory has the potential of further facilitating successful language teaching. General implications of the study This study aimed to test the applicability of CL theory to the instruction of conditionals within a mixed-methods design, using tasks as pedagogical means supporting the delivery of instructional content. Both treatment conditions—cognitive and task-supported—were effective for producing L2 development of the target form, as measured by the tests. In line with findings of DeKeyser and Prieto Botana’s (2015) meta-analysis supporting the benefit of explicit instruction for structures of greater complexity, this study demonstrated that explicit focus on form was effective for the classroom instruction of conditionals. It is important to discuss the interaction between CL instructional approach and raising metalinguistic awareness. Recent research by Roehr-Brackin (2014) suggests that explicit knowledge and learner’s reliance on metalinguistic awareness strategies can be a powerful tool for promoting L2 learning and use. If adapted to learners’ level, CL provides a framework for raising metalinguistic awareness systematically, which is a clear benefit of applying CLs in L2 instruction. Furthermore, the combination of the cognitive and task-supported instruction was more effective than the task-supported instruction alone. Pedagogic tasks proved to be an appropriate method for exploring authentic usage contexts. CLs appears to provide a better theory of language in general than the traditional (formal and structural) approaches to language analysis. Being a usage-based theory of language, CLs is highly compatible with research on task-based teaching, which also emphasizes language learning through meaningful content and contextualized language use. Task-supported instruction can function as a pedagogical platform complementary to the process of adapting the CL theory to language classroom. Combined with qualitative findings (see Jacobsen 2015 ), these results suggest that the focus on usage-based patterns and meaning indeed made a difference in the subjects’ performance. While pedagogic tasks are a valuable addition to the instructional process, the improved teaching methods alone are not sufficient. A more accurate and insightful theory of language is needed to make language teaching more meaningful overall. Limitations and future directions The study has a number of limitations. First of all, with N = 57, the sample size was relatively small, preventing the results from becoming more generalizable. It is also possible that some of the findings can be attributed to the context of this particular graduate EAP program as well as to the educational and cultural backgrounds of the majority participants. Replication with students from other backgrounds (i.e. taking into account other prior instructional experiences) would be instrumental for establishing the degree of generalizability of this study’s findings. Furthermore, echoing previously documented limitations ( Boers 2013 ), it was not possible to account for learners’ individual differences; additional data on learners’ motivation, working memory, and general aptitude (as defined by specific contexts) would need to be collected to determine the effects of possible intervening variables. Also, the whole instructional treatment lasted three weeks, but given that CL methods are relatively novel in classroom instruction, it would be necessary to test the long-term effects of a similar treatment. Establishing more specific effects of longitudinal and systematic classroom exposure to CL theory had previously been proposed as an area of recommended inquiry by other researchers ( Holme 2012 ; Tyler 2012 ). Furthermore, teacher buy-in and appropriate training would be crucial for successful replication of applied CL research. Another limitation of this study is such that the CL materials can only be used by individuals who both received prior training in CL and believe in its usefulness as a paradigm of language teaching. In a way, promoting CL as a basis for an L2 instructional system would require significant popularization of CL in teacher training departments and a CL-oriented shift in language teacher training curriculum. Until CL becomes a mainstream part of language teaching, we could not adopt directional hypotheses for L2 studies involving CL-informed methodology. In general, while this study does provide support for using adapted CL theory in the classroom, the research conducted thus far is not sufficient for establishing CL theory’s superiority over traditional instruction; additional future research needs to examine how CL theory in combination with systematic pedagogical interventions can be used for teaching other grammatical structures. Notes Natalia Dolgova Jacobsen is Teaching Assistant Professor of English for Academic Purposes at the George Washington University in Washington, DC. Her research interests include applied cognitive and corpus linguistics, task-based language teaching, second language writing in professional contexts, and English for Academic/Specific Purposes. Address for correspondence : English for Academic Purposes (EAP) Program, George Washington University, 2114 G Street NW, Washington, DC 20052, USA. < natalia@gwu.edu > 1 This study uses the task-supported model because of local context constraints (see ‘Methods’ section for details); however, TBLT is at the ideological root of task-supported instruction and is therefore addressed in this section. 2 The term ‘predictive’ refers to any expression of prediction conditioned ‘upon some unrealized event’ ( Dancygier and Sweetser 2005 : 28). If we use the traditional ESL terminology, all three conditional types—factual, hypothetical, and counterfactual—were addressed in the study. References Achard M. 2004 . ‘Grammatical instruction in the natural approach: A cognitive grammar view’ in Achard M. Niemeier S. (eds): Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching . Mouton de Gruyter . Achard M. Niemeier S. (eds). 2004 . Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching . Mouton de Gruyter . Boers F. 2013 . ‘Cognitive Semantic ways of teaching figurative phrases. An assessment’ in Gonzálvez-García F. Peña Cervel M. S. Pérez Hernández L. (eds): Metaphor and Metonymy Revisited Beyond the Contemporary Theory of Metaphor . John Benjamins Boers F. Lindstromberg S. . 2008 . Cognitive Linguistic Approaches to Teaching Vocabulary and Phraseology . Mouton de Gruyter . Bygate M. Skehan P. Swain M. (eds). 2001 . Researching Pedagogic Tasks, Second Language Learning, Teaching and Testing . Longman . Cadierno T. 2008 . ‘ Learning to talk about motion ’ in Robinson P. Ellis N. (eds): Handbook of Cognitive Linguistics and Second Language Acquisition . Routledge . Cadierno T. Robinson P. . 2009 . ‘Language typology, task complexity and the development of L2 lexicalization patterns for describing motion events ,’ AnnualReview of Cognitive Linguistics 6 : 245 – 77 . Google Scholar CrossRef Search ADS Celce-Murcia M. Larsen-Freeman D. . 1999 . The Grammar Book: An ESL/EFL Teacher’s Course . Heinle & Heinle . Comrie B. 1986 . ‘Conditionals: a typology’ in Traugott E. ter Meulen A. Snitzer Reilly J. Ferguson C. (eds): On conditionals . Cambridge University Press . Dancygier B. Sweetser E. . 2005 . Mental Spaces in Grammar: Conditional Constructions . Cambridge University Press . DeKeyser R. Prieto Botana G. . 2015 . ‘ The effectiveness of processing instruction in L2 grammar acquisition: A narrative review ,’ Applied Linguistics 36 : 290 – 305 . Google Scholar CrossRef Search ADS Dirven R. 2001 . ‘English phrasal verbs: theory and didactic application’ in Putz M. Niemeier S. Dirven R. (eds): Applied cognitive linguistics II: Language Pedagogy . Mouton de Gruyter . Eckerth J. 2008a . ‘ Investigating consciousness-raising tasks: Pedagogically-targeted and non-targeted learning gains ,’ International Journal of Applied Linguistics 18 : 119 – 45 . Google Scholar CrossRef Search ADS Eckerth J. 2008b . ‘ Task-based learner-learner interaction ’ in Eckerth J. Siekmann S. (eds): Task-based Language Learning and Teaching: Theoretical, Methodological, and Pedagogical Perspectives . Peter Lang . Ellis N. 2002 . ‘ Frequency effects in language processing: A review with implications for theories of implicit and explicit language acquisition ,’ Studies in Second Language Acquisition 24 : 143 – 88 . Ellis R. 2003 . Task-based Language Learning and Teaching . Oxford University Press . Ellis R. 2004 . ‘ The definition and measurement of L2 explicit knowledge ,’ Language Learning 54 : 227 – 75 . Google Scholar CrossRef Search ADS Ellis R. 2009 . ‘ Task-based language teaching: Sorting out the misunderstandings ,’ International Journal of Applied Linguistics 19 : 222 – 46 . Google Scholar CrossRef Search ADS Evans J. Over D. . 2004 . If . Oxford University Press . Fauconnier G. Turner M. . 2002 . The Way We Think: Conceptual Blending and the Mind's Hidden Complexities . Basic Books . Fordyce K. 2014 . ‘ The differential effects of explicit and implicit instruction on EFL learners’ use of epistemic stance ,’ Applied Linguistics 35 : 6 – 28 . Google Scholar CrossRef Search ADS Fotos S. 1993 . ‘ Consciousness raising and noticing through focus on form: Grammar task performance versus formal instruction ,’ Applied Linguistics 14 : 385 – 407 . Google Scholar CrossRef Search ADS Fotos S. 1994 . ‘ Integrating grammar instruction and communicative language use through grammar consciousness-raising tasks ,’ TESOL Quarterly 28 : 323 – 51 . Google Scholar CrossRef Search ADS Fotos S. Ellis R. . 1991 . ‘ Communicating about grammar: A task-based approach ,’ TESOL Quarterly 25 : 605 – 28 . Google Scholar CrossRef Search ADS De la Fuente M. 2006 . ‘ Classroom L2 vocabulary acquisition: Investigating the role of pedagogical tasks and form-focused instruction ,’ Language Teaching Research 10 : 263 – 95 . Google Scholar CrossRef Search ADS Han Y. Ellis R. . 1998 . ‘ Implicit knowledge, explicit knowledge and general language proficiency ,’ Language Teaching Research 2 : 1 – 23 . Google Scholar CrossRef Search ADS Holme R. 2009 . Cognitive Linguistics and Language Teaching . Palgrave MacMillan . Holme R. 2012 . ‘ Cognitive linguistics and the second language classroom ,’ TESOL Quarterly 46 : 6 – 29 . Google Scholar CrossRef Search ADS Hulstijn J. H. de Graaff R. . 1994 . ‘ Under what conditions does explicit knowledge of a second language facilitate the acquisition of implicit knowledge? A research proposal ,’ AILA Review 11 : 97 – 112 . Jackson F. 1991 . Conditionals . Oxford University Press . Jacobsen N. D. 2012 . Applying Cognitive Linguistics and Task-supported Language Teaching to Instruction of English Conditional Phrases . Washington, DC: Georgetown University , Unpublished Dissertation. Jacobsen N. 2015 . ‘A cognitive linguistic analysis of English conditionals in English for Academic Purposes (EAP) instruction: Implications from sociocultural theory’ in Masuda K. Arnett C. Labarca A. (eds): Cognitive Linguistics and Sociocultural Theory: Applications for Second and Foreign Language Teaching . Mouton de Gruyter . Langacker R. W. 2008 . ‘Cognitive grammar as a basis for language instruction’ in Robinson P. Ellis N. (eds): Handbook of Cognitive Linguistics and Second Language Acquisition . Routledge . Lightbown P. 1985 . ‘ Great expectations: Second language acquisition research and classroom teaching ,’ Applied Linguistics 6 : 173 – 89 . Google Scholar CrossRef Search ADS Lightbown P. 2000 . ‘ Anniversary article: Classroom SLA research and second language teaching ,’ Applied Linguistics 21 : 431 – 62 . Google Scholar CrossRef Search ADS Liu D. 2010 . ‘ Going beyond patterns: Involving cognitive analysis in the learning of collocations ,’ TESOL Quarterly 44 : 4 – 30 . Google Scholar CrossRef Search ADS Lindstromberg S. 1996 . ‘ Prepositions: meaning and method ,’ ELT Journal 50 : 225 – 36 . Google Scholar CrossRef Search ADS Lindstromberg S. Boers F. . 2005 . ‘ From movement to metaphor with manner-of-movement verbs ,’ Applied Linguistics 26 : 241 – 61 . Google Scholar CrossRef Search ADS Littlemore J. Low G. . 2006 . ‘ Metaphoric competence, second language learning, and communicative language ability ,’ Applied Linguistics 27 : 268 – 94 . Google Scholar CrossRef Search ADS Long M. 2000 . ‘Focus on form in task-based language teaching’ in Lambert R. Shohamy E. (eds): Language Policy and Pedagogy: Essays in Honor of A. Ronald Walton . John Benjamins . Meunier F. 2008 . ‘Corpora, cognition and pedagogical grammars’ in de Knop S. D. de Rycker T. D. (eds): Cognitive Approaches to Pedagogical Grammar: A Volume In honour of Rene Dirven . Mouton de Gruyter . Moder C. L. 2010 . ‘To learn effectively: Grammatical constructions for second language grammar instruction’. Paper presented at AAAL, Atlanta, GA. Niemeier S. Reif M. . 2008 . ‘Applying cognitive grammar to tense-aspect teaching’ in de Knop S. de Rycker T. (eds): Cognitive Approaches to Pedagogical Grammar: A Volume In honour of Rene Dirven . Mouton de Gruyter . Norris J. Ortega L. . 2000 . ‘ Effectiveness of L2 instruction: A research synthesis and quantitative meta-analysis ,’ Language Learning 50 : 417 – 528 . Google Scholar CrossRef Search ADS Révész A. 2009 . ‘ Task complexity, focus on form, and second language development ,’ Studies in Second Language Acquisition 31 :/ 3 : 437 – 470 . Google Scholar CrossRef Search ADS Robinson P. 2001 . ‘Task complexity, cognitive resources, and syllabus design’ in Robinson P. (ed.): Cognition and Second Language Instruction . Cambridge University Press . Robinson P. 2005 . ‘ Cognitive complexity and task sequencing ,’ International Review of Applied Linguistics 43 : 1 – 32 . Google Scholar CrossRef Search ADS Robinson P. 2011 . ‘ Task-based language learning: A review of issues ,’ Language Learning 61(Supplement 1) : 1 – 36 . Google Scholar CrossRef Search ADS Robinson P. Ellis N. (eds.). 2008 . Handbook of Cognitive Linguistics and Second Language Acquisition . Routledge . Roehr-Brackin K. 2014 . ‘ Explicit knowledge and processes from a usage-based perspective: The developmental trajectory of an instructed L2 learner ,’ Language Learning 64 : 771 – 808 . Google Scholar CrossRef Search ADS Schmidt R. 2001 . ‘Attention’ in Robinson P. (ed.): Cognition and Second Language Instruction . Cambridge University Press . Samuda V. Bygate M. . 2008 . Tasks in Second Language Learning . Palgrave Macmillan . Sharwood Smith M. 1981 . ‘ Consciousness-raising and the second language learner ,’ Applied Linguistics 2 / 2 : 159 – 68 . Google Scholar CrossRef Search ADS Sharwood Smith M. 1993 . ‘ Input enhancement in instructed SLA: Theoretical bases ,’ Studies in Second Language Acquisition 15 : 165 – 79 . Google Scholar CrossRef Search ADS Sharwood Smith M. 2007 . ‘Morphological and syntactic awareness in foreign/second language learning’ in Cenoz J. Hornberger N. (eds): Encyclopedia of Language and Education . Kluwer Academic Publishers . Spada N. Tomita Y. . 2010 . ‘ Interactions between type of instruction and type of language feature: A meta-analysis ,’ Language Learning 60 : 263 – 308 . Google Scholar CrossRef Search ADS Swales J. Feak C. . 1998 . Academic Writing for Graduate Students. University of Michigan Press . Taylor J. 1997 . ‘Conditionals and polarity’ in Athanasiadou A. Dirven R. (eds): On Conditionals Again . John Benjamins . Tyler A. 2012 . Cognitive Linguistics and Second Language Learning: Theoretical Basics and Experimental Evidence . Routledge Tyler A. Evans V. . 2004 . ‘ Applying cognitive linguistics to pedagogical grammar: The case of over ’ in Achard M. Niemeier S. (eds): Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching . Mouton de Gruyter . Tyler A. Mueller C. Ho V. . 2011 . ‘ Applying cognitive linguistics to learning the semantics of English to, for, and at: An experimental investigation ,’ Vigo International Journal of Applied Linguistics 8 : 181 – 205 . Van den Branden K. Bygate M. Norris J. . (eds) 2009 . Task-based Language Teaching: A Reader . John Benjamins . Werth P. 1997 . ‘Conditionality as cognitive distance’ in Athanasiadou A. Dirven R. (eds): On Conditionals Again . John Benjamins . Woods M. 1997 . Conditionals . Clarendon Press . Yasuda S. 2010 . ‘ Learning phrasal verbs through conceptual metaphors: A case of Japanese EFL learners ,’ TESOL Quarterly 44 : 250 – 73 . Google Scholar CrossRef Search ADS © Oxford University Press 2016

Journal

Applied LinguisticsOxford University Press

Published: Sep 28, 2016

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off