An SLA-Informed and Cognitive Linguistic Approach to the Teaching of L2 Catalan Tense-Aspect

This article explores the relationship between second language acquisition and language teaching presenting the design of a teaching sequence on past tenses that integrates the results of research on tense and aspect in a second language (the Aspect and Discourse Hypotheses and Input processing instruction). By implementing the principles of cognitive linguistics, the article presents the design of activities to teach the meanings of perfective, imperfective, and perfect morphology with the introduction of the notions of space of action, verbal action, and the speaker's point of view. All concepts are illustrated with activities implemented in a Catalan classroom at the University of Cardiff (A1 and A2 levels).


Introduction
The teaching of L2 tense-aspect is prominent in the L2 teaching of Romance and non-Romance languages, mostly because different languages encode temporality in different ways and L2 adult learners have a hard time learning the different mechanisms to express temporality. There are a number of second language acquisition (SLA) studies that investigate the acquisition of L2 tense-aspect and provide a rich description of how L2 tense and aspect forms and meanings are learned (Bardovi-Harlig, 2000;Bardovi-Harlig & Comajoan-Colomé, 2020; Comajoan-Colomé, 1 Universitat de Vic -Universitat Central de Catalunya. Email: llorenc.comajoan@uvic.cat 2 Fitzwilliam College -University of Cambridge. Email: al2027@cam.ac.uk MONOGRÁFICO Círculo de Lingüística Aplicada a la Comunicación ISSN: 1576-4737 https://dx.doi.org /10.5209/clac.76714 2014; Salaberry, 2008). Even though one would imagine that SLA research was followed by classroom studies to investigate the effectiveness and practicality of research findings, this has not been the case. The reasons for such a lack of relationship between theory and practice in language teaching are twofold. First, SLA theory does not always translate well to classroom teaching. For instance, SLA studies that focus on syntax tend to focus on semantic and morphological features that are not readily extended to classroom practices. Second, results from SLA theory are often inconclusive in the sense that results from one study do not always justify applying them to specific classroom practices. This article attempts to fill the gap between SLA theory and classroom language teaching by describing the design and implementation of a sequence to teach L2 Catalan tense-aspect meanings in the classroom through a cognitive linguistic perspective. The proposal integrates findings from SLA research and presents a description of teaching activities that can be applied to other Romance languages with similar tense-aspect systems. The article is divided into five sections. In the first one, the relationship between theory and practice in language learning and teaching is explored. The second section is devoted to the acquisition and teaching of tense-aspect in L2 Catalan. In the third section, a proposal for the teaching of L2 tense-aspect in the foreign language classroom based on SLA findings and cognitive linguistics (CL) is presented. The following section discusses the advantages and drawbacks of designing a teaching sequence informed by SLA findings and CL. Conclusions are presented in section five.

SLA findings and teaching implications
The relationship between theory and practice in language teaching has traditionally been problematic as has often been the case in other disciplines as well (cf. the debate between basic and applied sciences). Ellis (2015), when discussing the application of SLA into language teaching practice, summarizes three main ways of linking research and teaching. First, there is the possibility of basing pedagogical proposals on the findings of SLA. This position is often related to the application of SLA research to classroom teaching. This is problematic when the findings are limited to one study, but it can be fruitful when the results from SLA are solid and, thus, relevant (Ortega, 2012) or consistent (Lightbown, 2000) to be adopted for classroom teaching. The second position -practitioner research and SLA -encourages teachers to carry out their own practitioner research in their classrooms by designing small-scale studies or doing replication studies. Even though this type of research is needed and desirable, as admitted by Ellis, it is time consuming and current studies have showed that teachers often do not have the skills or time to carry it out. Finally, Ellis advocates for the third position, one of exploring language pedagogy through SLA; that is, detecting some problematic issues in classroom teaching and examining them through the lens of SLA research to study "the extent to which current pedagogic views are supported or challenged by SLA research" (p. 311).
The positions described above are related to several issues that make the connection between theory and practice complex. Thus, in recent studies, researchers focusing on the connection between theory and practice have documented that language teachers are indeed aware of research and interested in conducting their own studies but, at the same time, they have found obstacles to carry out research, e.g., lack of training, no professional recognition (lack of time), and lack of support from their institutions (Marsden & Kasprowicz, 2017;Muñoz-Basols, Rodríguez-Lifante, & Cruz-Moya, 2017). One possible solution to some of these constraints is for researchers to be aware of them and devote research to matters relevant to teachers and provide easier access. In this respect, recent initiatives such as the publication of practical summaries by SLA and language teaching journals and the current interest in research with implications for the classroom are a welcome addition (DeKeyser & Botana, 2019;Marsden, Trofimovich, & Ellis, 2019;Sato & Loewen, 2019).
The acquisition of L2 tense-aspect figures prominently as one of the most researched topics in SLA. Even though different theories and methodological approaches have attempted to explain the acquisition of tense-aspect (see Comajoan-Colomé, 2014 for a review on L2 Spanish), the most well-researched theoretical accounts that are amenable to classroom teaching are the so-called "Aspect Hypothesis" and "Discourse Hypothesis" (Bardovi-Harlig, 2000), because both focus on the development of learner interlanguage. A third theory, Input Processing, also has several implications for classroom practice and has been tested in studies dealing with the acquisition of tense-aspect (Van Patten, 2010Shintani, 2015).
Briefly, the Aspect Hypothesis states that the acquisition of past morphology is mostly influenced (in the beginning states of acquisition) by the lexical aspect of verbal predicates: L2 perfective morphology emerges earlier and in telic predicates (those that have an inherent endpoint, e.g., make a cake, arrive home at 10) and later extends to atelic predicates (without endpoints, e.g., be interesting, want to dance). In contrast, imperfective morphology emerges later and follows the opposite path, emerging in atelic predicates and later extending to telic ones (Bardovi-Harlig, 2000). The Discourse Hypothesis states that L2 learners use past morphology in narratives to build discourse grounding in a predictable manner: in the initial stages of acquisition, learners build narratives mostly with foreground, which is marked with perfective morphology; and with time, they build narratives that combine foreground (mostly with perfective forms) and background (with emerging and developing imperfective forms) ( Bardovi-Harlig, 2000). Finally, Input Processing, unlike the other two hypotheses, is a general learning theory and does not exclusively focus on past tense. It is defined as a type of pedagogical intervention focusing on form, which attempts to describe the constraints of the form-meaning relationship when processed by learners and the pedagogical ways to circumvent such constraints (Van Patten, 2017).
As opposed to the number of studies on the acquisition of tense-aspect, the number of classroom intervention studies that investigate the acquisition of tense-aspect is scarce (Bielak & Pawlak, 2013;Kermer, 2016;Niemeier, 2017). Some research does not include an empirical study on how to teach tense-aspect but discusses how SLA findings on the acquisition of tense-aspect can be related to classroom teaching. For instance, Blyth (2005) discussed the acquisition of tense-aspect in Spanish, arguing that it is a highly complex phenomenon because of the confluence of several issues. In his discussion of SLA research on the acquisition of tense-aspect and language pedagogy, he argued for "three organizing principles to guide teachers' classroom practice" ( Table 1). The principles discussed by Blyth are applied to the design of a teaching sequence for the teaching of L2 Catalan in Section 3 of this article. Table 1. Organizing principles to guide teachers' classroom practice (compiled from Blyth, 2005) Findings from SLA empirical studies Principles to guide teachers' classroom practice 1. Learners acquire aspectual distinctions in gradual, developmental stages that reflect prototypical associations between lexical aspect, discourse grounding and grammatical form.
1. Design pedagogical interventions to enhance the input in keeping with students' developmental readiness.
2. Aspect relates to viewpoint or perspective-taking, which in turn is grounded in visual perception.

2.
Base grammatical explanations and activities as much as possible on the students' own visual perception of events.
3. Aspectual usage is highly sensitive to both task and genre. 3. Choose appropriate narrative texts and tasks that take into account cognitive and linguistic complexity as well as native speaker norms.
Llopis-García, Real Espinosa and Ruiz Campillo (2012) critically evaluated the different ways (or methods) in whi11ch L2 Spanish tense-aspect is taught in the classroom and listed three main ones: (a) taxonomic: based on the provision of lists of functions for specific verbal forms; (b) formalist: based on the association of markers (mostly adverbial expressions) for each tense-aspect form; and (c) discursive-functionalist: based on the grounding characteristics of narratives (e.g., action or perfective past in the foreground; description or imperfective past in the background). In opposition to the three approaches, Llopis-García et al. (2012) argued for a fourth approach based on CL. Such an approach is the one discussed and applied into the teaching of L2 Catalan tense-aspect in Section 3.
In addition to the approaches discussed by Llopis-García et al. (2012), two more can be added. First, Input Processing, which is based on designing input processing activities that make learners focus on particular features of tense and aspect and foster appropriate form-meaning connections; e.g., by suppressing all adverbials from a list of sentences and having learners process whether a sentence refers to the present or the past (see Benati & Lee, 2008;Alonso-Aparicio & Llopis-Garcia, 2019 for specific examples on input processing activities to teach tense-aspect). Second, Concept-based Instruction, which follows social constructivist theory and focuses on the learners' metalinguistic reflections and linguistic justifications for the learning of tense-aspect. In this approach, classroom materials include learners' thoughts about the use of tense-aspect features and foster collaboration among learners (Negueruela & Lantolf, 2006).

Cognitive linguistics (CL) and language teaching
CL is a usage-based theory of language developed in the late 1980s that emphasizes that language reflects general cognitive processes. Tyler (2012), examining the connection between CL and language teaching, described it as an approach with the following guiding principles: (a) there is no sharp distinction between the lexicon and the grammar; (b) meaning is grounded in our everyday interactions with the world around us and the nature of our bodies; (c) linguistic units constitute complex categories (not only based on taxonomic connections but also on metaphoric and metonymic extensions, with both prototypical and non-prototypical members); and (d) language is usage-based (see also Niemeier, 2017). With the development of CL and its applicability to language teaching emerged applied cognitive linguistics, the field "concerned with the acquisitional and pedagogical implications of Cognitive Linguistics in Second and Foreign Language Teaching/Learning" (Pütz, 2007(Pütz, , p. 1139. The adoption of CL in language teaching presents two main advantages (Achard, 2008, p. 451). First, a focus on the symbolic nature of all language expressions (lexical and grammatical) is congruent with current language teaching that emphasizes the role of meaning in language learning. Furthermore, such an approach allows for the teaching of grammar that is compatible with communicative approaches to language teaching. Second, CL emphasizes that the distribution of language expressions depends on the choices made by speakers and not so much on the language properties per se and that such a principle is fully compatible with the idea that language learners are at the center of the communicative act (cf. communicative and task-based approaches to language teaching; Niemeier, 2017).
The main tenets of CL, as applied to L2 grammar teaching, contrast sharply with the so-called traditional (i.e., based on structural linguistics) approach to language teaching, as summarized in Table 2 (from Bielak & Pawlak, 2013, p. 128). Discourse and other pragmatic factors not/rarely included in the meaning of grammatical elements The meanings of declarative clauses do not depend on the meaning of neighbor sentences

Grammar is motivated
The use of the nonprogressive present tense with performative clearly explained by the semantics of these elements Grammar is often arbitrary The use of the non-progressive present tense with performatives not explained 6. Little/no vagueness and imprecision of description ---Significant vagueness and imprecision of description Hedging of rules by words such as normally and usually, not followed by reference to "abnormal" or "unusual" cases 7. No contradiction between different rules/sub-rules ---Contradiction between rules/ sub-rules The rule that the non-progressive present is used with such (dynamic) verbs as promise and refuse followed by the rule that the nonprogressive present refers to general permanent characteristics of people and things. Despite the usual positive promises when a new approach comes into the arena, with time, different positions on the potential of CL for language pedagogy emerged from most to less optimistic: Although CL is a relatively new framework, it is now sufficiently developed to be usefully applied to language learning. Indeed, several researchers have begun to apply CL to Language pedagogy. (Tyler, 2012, p. 25, our emphasis) It may be unlikely that cognitive linguistics will result in a radically new teaching methodology both in the specific case of pedagogical grammar and in language teaching in general. Rather, cognitive models of language highlight certain aspects of already existing language teaching methodologies which deserve further attention. (Broccias, 2008, p. 87, our emphasis) In addition, a major pitfall of earlier cognitive accounts of language teaching was the lack of specific empirical support for the new approach. However, in the second decade of the 21 st century, several studies began to examine the effectiveness of the cognitive approach to language teaching. In this section, we focus on those studies that specifically studied the acquisition of L2 tense-aspect (Bielak & Pawlak, 2013;Kermer, 2016Kermer, , 2019Niemeier & Reif, 2008;Niemeier, 2017;and Tyler, 2012). Bielak and Pawlak (2013) compared the traditional and cognitive approaches to the teaching of L2 English in the acquisition of present and progressive forms by senior high school students in Poland and found that both approaches were similarly effective (compared to a control group). Interestingly, the study contained a questionnaire about the attitudes of learners toward the two approaches and, in general, learners found the cognitive approach more complex and displayed a less positive attitude toward the traditional approach. Kermer (2016) conducted two studies that addressed the comparison between a traditional approach and a cognitive one. In the first study, she investigated the acquisition of English by German young learners (average age: 11) who were learning the distinction between simple present and present progressive and found mixed results. For some tasks, the cognitive approach was superior (e.g., in receptive knowledge), but in other tasks, the traditional one was superior to the cognitive approach. In the second study, she investigated the acquisition of the contrast between the present perfect and simple past by German learners of L2 English (age rank: [12][13][14]. The results showed that the traditional group obtained better results than the cognitive one. Finally, Alonso-Aparicio and Llopis-García (2019) found that a cognitive approach to tense-aspect in L2 Spanish (preterite vs. imperfect) resulted in better results than a traditional approach (based on adverbials and lists of functions) in production but not in interpretation data. In sum, the general picture of the SLA studies that have compared traditional and cognitive approaches is that the cognitive perspective produces mixed results in the sense that in most studies the cognitive approach is at least as beneficial as the traditional approach. The reasons for the mixed results are related to issues such as those mentioned by Bielak & Pawlak (2013): (a) cognitive principles may be too complex, compared to the simplicity of the so-called traditional rules; (b) cognitive approaches may be more suitable for higher level learners; and (c) it is difficult to combine a cognitive approach, which often focuses on grammar, with a communicative approach.
Considering the results from SLA research focusing on the role of lexical aspect and discourse grounding and current developments of applied cognitive linguistics, this paper presents the design of a teaching sequence that addresses some of the pitfalls documented in previous studies. It targets complex principles related to three tenses and their associated meanings, instead of targeting two tenses as has been previously done. The teaching sequence based on CL is implemented in a beginner group and it is embedded in communicative language teaching. The goal is to provide an instance of the link between SLA research and language teaching by examining a problematic feature for language teachers (i.e., the contrast between perfective, imperfective, and perfect past morphology) and attempting to show how research can be relevant for the design of teaching sequences for the classroom. The design and implementation of the teaching sequence is also an example of collaboration between a pre-service teacher and linguist (Llop Naya) and a teacher and SLA researcher (Comajoan-Colomé). Both identified a problematic area of L2 language teaching (tense-aspect) while Llop Naya was an M.A. student and a pre-service teacher in an SLA course taught by Comajoan-Colomé and later collaborated in the design of the teaching sequence, which was implemented by Llop in her own classroom. Even though the project was not designed as an action-research project, it shares some of its characteristics, such as the focus on developing autonomous professional judgments, fostering the feeling of understanding the complexity of classroom teaching and reducing the feeling of frustration, and allowing for teacher-researcher collaboration (Burns, 2019).

The acquisition and teaching of tense-aspect in L2 Catalan
The difficulty of teaching and learning the past tense-aspect system in Catalan has to do with distinguishing between the use of not two (as in other Romance languages) but three forms (see Pérez Saldanya, 2002, for a more detailed description of the past tense in Catalan): passat perifràstic (PP), imperfet, and perfet. The English terms we use in this article to refer to the three forms are PP (periphrastic past, preterite), imperfect, and perfect.
PP forms are perfective and are used for actions that are considered finished and bounded by the speaker, e.g., Ahir vaig comprar un cotxe nou 'Yesterday, I bought a new car'. The PP in Catalan is formed by a compound form of a grammaticalized form of the verb anar, 'to go', followed by the infinitive, a feature that can be easily confounded by learners for a future. Catalan does not have a periphrastic future with go + infinitive, but a synthetic one; e.g., Demà no aniré a la feina, 'Tomorrow I will not go / am not going to go to work'. The Catalan PP corresponds to the perfective meaning of the so-called preterite, passé composé, or passato prossimo in Spanish, French, and Italian, respectively.
Catalan imperfect forms encode actions that are viewed by the speaker as open; hence, their progressive, continuous, or habitual meanings: Quan el sol es ponia… 'When the sun was setting down'; La ciutat era sorollosa 'The city was noisy'; Quan era jove, jugava a futbol 'When I was younger, I used to play soccer'.
Catalan perfect forms encode the meaning of current relevance, and they have three basic meanings: resultative (M'he trencat la cama 'I have broken my leg'), experiential (He estat a París 'I have been to Paris'), and inclusive (He estudiat arquitectura 'I have studied Architecture'), all of them with corresponding meanings in Romance languages.
The use of the three Catalan tenses is equivalent to the use of similar forms in other Romance languages (e.g., standard Peninsular Spanish), except for one feature. Temporal distance is marked obligatorily in the past in Catalan. Thus, for hodiernal (within the same day of the utterance moment) perfective meanings, it is obligatory to use the perfect in Catalan, whereas in other Romance languages, perfective and perfect forms may be used. For instance, in (1a), in a sentence in the past with avui 'today', Catalan uses obligatorily a perfect form, whereas other Romance languages show regional variation (e.g., preterite and perfect in Spanish) or obligatorily use a form that coincides with perfect (e.g., French passé compose and Italian passato prossimo). In contrast, with prehodiernal meanings, e.g., ahir 'yesterday', Catalan uses PP forms: (1a) Avui he comprat el diari (perfect form).
Research in the acquisition of L2 Catalan tense-aspect has documented that current teaching materials (i.e., course textbooks) are eclectic in the sense that they combine different approaches to the teaching of tense-aspect (e.g., list of functions, focus on adverbials, and rules of thumb), but none of them incorporate CL or SLA findings in an explicit manner (Comajoan-Colomé, 2015). Furthermore, there is heterogeneity in how L2 Catalan teaching materials introduce past-tense forms and meanings for the first time. While some textbooks begin with the imperfect (habitual meaning), followed by the PP (finished actions) and the perfect, other materials begin with the PP, followed by the imperfect and the perfect. The reasons for such disparity in the order in which past tenses are taught are not clear, but they could be related to the perception by teachers that the PP, which is a compound form in Catalan and can easily be confounded with a future, is more difficult than the imperfect and thus it is taught first. On the other hand, authors of teaching materials who decide to introduce the PP first may understand that its core meaning (finished pasts action) is easier to process than that of the imperfect (unboundedness). In order to solve this sequencing problem, the current proposal begins with the perfective meaning, followed by the imperfect and the perfect. The reason behind this is that consistent findings from SLA research have found that perfective markers are easier to learn for learners than imperfective ones due to the cognitive salience of perfectivity, as opposed to imperfectivity and perfect (Bardovi-Harlig, 2000;Comajoan-Colomé, 2014;Salaberry, 2008).

An SLA-principled approach
This section presents the teaching sequence that was designed and implemented for the teaching of L2 past tenses in Catalan. The design of the sequence is SLA-principled in the sense that it incorporates previous research on CL and Blyth's (2005) SLA principles to guide teachers' classroom practice in the teaching of tense-aspect (Table 1). For the first principle, regarding the design of classroom interventions that enhance input according to the students' developmental readiness, the sequence incorporated input processing activities that asked the learners to process language input for complex meanings before asking them to produce new language output. Furthermore, the input for activities was selected according to SLA research findings regarding the emergence of inflection of past in specific verbal predicates (e.g., lexical-aspectual characteristics of predicates) and parts of discourse (narrative grounding).
For the second principle, focusing on providing grammatical explanations and activities that consider the students' visual perception, we analyzed the conceptual and linguistic notions that could be adapted from CL. The sequence was based on the principles of CL transformed into "operative principles in the target language that will enable them [learners] to control their performance without the sheer memorization of random lists of notions and functions" (Llopis-García, 2016, p. 32). Bearing the cognitive approach in mind, a prototype for each of the three past verbal tenses was prepared to be introduced in the classroom. The proposal intended to equip students with a simpler and more unitary explanation of past tense usage than in traditional teaching materials (often loaded with rules and exceptions) to enhance their comprehension of the prototypical meaning of the three main past verbal tenses in Catalan. Following Blyth's recommendations (2005), illustrations of meaning were crucial to the design of the sequence. In this respect, the sequence explores how graphic and audio-visual image-based reasoning (as a metaphorical projection to capture abstract domains) can be a useful tool to delve into how grammar phenomena function (Lakoff & Johnson, 1980;Johnson, 1987;Lakoff, 1993).
To present the prototypes of each tense, we adopted the cognitive linguistic perspective that tense-aspect can be described through three main concepts (Llopis-García, 2016): (a) the space of the action (tense), which indicates whether the event takes place or not in an actual space (here 'now' vs. there 'then'); (b) the verbal action (aspect); that is, if the action is a completed process or not (perfective and perfect aspect vs. imperfective); and (c) the point of view of the speaker or their perspective (aspect), depending on whether the observer is considered to be "within" the action (internal observer), and thus to partake it, or not (external observer). The distinction between the verbal action and the point of view of the speaker are clearly related notions, as the boundedness or unboundedness of the verbal action of each tense implies the perspective of the speakers to be external or internal. The notion of point of view of the speaker is used here not as an alternative conceptualization of the grammatical notion of aspect, but as an alternative (subjective) pedagogical tool that makes speakers aware of the impact of the nature of the verbal action (or the perfective/imperfective aspectual distinction) on their role as conceptualizers of the event.
In order to introduce the three concepts that characterize the cognitive linguistic approach to tense-aspect, a number of PowerPoint slides designed by the instructor of the course (Llop Naya) were prepared (Figures 1-3). The graphic representations of the prototypes were inspired by those in Ruiz Campillo (2014) and Alonso-Aparicio and Llopis-García (2019). Figure 1 shows the space of the action of the three Catalan past tenses and marks in red the different spaces of action for, on the one hand, PP and imperfect (there; that is, past) and, on the other hand, perfect (here; that is, current relevance). The slides had the purpose of introducing the prototypical meanings of the three tenses and were not intended to cover all meanings for each tense, since this was the first presentation of past tenses for A1-level students. For instance, Figure 1 shows that the imperfect is related to the past (there), even though it can refer to current events as well, when combined with a present perfect (e.g., He tornat a posar la carn al forn perquè estava (imperfect) crua; 'I have put the meat back in the oven because it was (imperfect) raw').   Figure 3 introduces the notion of the observer of the action (aspect) and marks in green the coincidence of the speaker being an internal observer of the action in the perfect and the imperfect, and the observer in the PP being external. For each tense, at the beginning of instruction, the prototype was presented and wherever necessary it was compared to the previous ones. Finally, for Blyth's (2005) third principle, regarding the selection of appropriate genres and activities to demonstrate aspectual usage, we adopted narrative texts as a basic cognitive experience shared by all humans as they try to understand experience (Berman, 2009;Thornborrow, 2012). In narratives, one can identify the foreground, which leads the storyline, and the background, which elaborates and makes comments on it (Bardovi-Harlig, 2000;Comajoan-Colomé, 2005). Narrative texts also foster the prototypical associations between lexical aspect, discourse grounding, and grammatical forms (Hopper, 1979;Reinhart, 1984;Bardovi-Harlig, 1994. This makes narrative texts optimal when teaching and learning Catalan past verbal tenses in a progressive and guided way. In addition, in the design of the activities, we adopted input processing principles; more specifically, the Lexical Preference principle (Van Patten, 2017, p. 169): the assumption that learners first process lexical items as cues for morphological features (e.g., relying upon adverbials for tense-aspect information rather than verbal morphology). Input processing activities in the teaching sequence did not include temporal markers in order to avoid the association of specific verbal tenses with temporal markers. Additionally, the sequence included an awareness-raising activity that focused on dissociating temporal markers from being used exclusively with one or another verbal tense. The absence of temporal markers only affected the activities of the teaching sequence oriented to input flooding (and processing). Additional materials used in the sequence, such as the reading comprehension activities or the assigned book (Pau Casals' biography), included temporal markers.
In the teaching sequence, we privileged primary exposure to input and, specifically, to biased input. Teaching the prototypical associations of verbal morphology, lexical aspect, and discourse grounding is based on SLA findings and instructed second language acquisition (ISLA) principles. SLA research has showed that L2 learners in the early stages of L2 acquisition tend to produce prototypical combinations of verbal morphology and lexico-aspectual characteristics of predicates in a predictable manner (e.g., perfective in telic predicates first, imperfective in telic predicates later) (see a review of the studies in Bardovi-Harlig & Comajoan-Colomé, 2020, but see also Domínguez, Tracy-Ventura, Arche, Mitchell, & Myles, 2013;González & Quintana Hernández, 2018;and Quintana Hernández, 2019). The reason for such pairings has to do with a cognitive universal tendency (i.e., the congruence of inflecting perfective salient actions with telic predicates) and frequency in the input of such pairings. In this respect, having students produce prototypical associations is warranted for two reasons. First, such pairings are easy to notice in the input at the early stages and thus may be more readily processed and produced by L2 learners. This is particularly important when the form-meaning pairings of past marking are not symmetrical in the L1 and the L2 (e.g., English past tense with -ed and past progressive with was/were -ing vs. Catalan past with past periphrastic, imperfect, and perfect). Second, having students notice and produce prototypical combinations makes it easier for them to cognitively process and understand the new form-meaning combinations. Eventually, learners will need to understand that all predicates may be inflected in all past tense forms (in prototypical and non-prototypical combinations), but in the early stages, providing all the possible combinations may cognitively overload learners.
ISLA has since its inception made the effort to link SLA and language teaching practice (Loewen, 2015). Regarding the teaching of grammar (e.g., past tense morphology and meanings), ISLA has focused on several facts about SLA that are closely related to teaching, such as the two following ones: (a) "Second language learners develop an implicit, abstract, and complex mental representation of language that bears no resemblance to rules in the traditional sense," and (b) "the development of mental representation for a second language has three characteristics: (1) it is slow; (2) it is piecemeal; (3) it is stage-like" (Van Patten, 2017, pp. 47-48; see also Salaberry, Comajoan-Colomé, & González, 2013). In this sense, even though from a purely linguistic perspective all verbal predicates may be inflected in the three Catalan past tenses (periphrastic past, imperfect, and prefect), that does not imply that all of them must be taught at the same time, since learners develop their own implicit representation, they do so in a stage-like manner, and they learn only what they are ready to learn (cf. Pienemann's 2015 processability theory). Blyth (2005) focused on the link between SLA findings and the teaching of tense-aspect and related it to what L2 teachers can expect from learners tackling the acquisition of L2 tense-aspect according to their interlanguage development. Regarding the expectations for learner production in the beginning stage, Blyth argued that "it seems reasonable that the goal of instruction should be to help learners overcome the past default marker [e.g., use preterite for all past meanings in Romance], and to acquire the prototypical uses of both the preterite and imperfect. Narrative input at this stage should have a clear foreground with telic events in the preterite and a few background clauses in the imperfect used mainly for scene-setting purposes (who, what, when, where)" (p. 221). And it is not until later (Blyth suggests the second or third year of college study) that non-prototypical uses should be introduced.
In sum, following Andersen and Shirai (1996), we reinforced the inner predisposition by learners to first acquire and use congruent combinations of form and meaning (see also Rastelli, 2020, p. 32). For this reason, prototypal associations of lexical aspect, discourse grounding, and verbal morphology were specially promoted in the early sessions for the Catalan A1 and A2 courses described in the following sections. Following Blyth's (2005) recommendations and Van Patten's (2017) facts about SLA, non-prototypical meanings were introduced progressively to help learners to understand that the three verbal tenses are compatible with all sorts of predicates (see Appendix 1).

Teaching sequence for the instruction of past tense-aspect in L2 Catalan
The teaching sequence was designed to be implemented in general courses of L2 Catalan for adult learners. The sequence was first designed as part of an M. A. Thesis in Catalan teaching by Ares Llop Naya and was later implemented in two teaching settings: at the Escola Oficial d'Idiomes (Official Language School) in Barcelona (Spain) and the School of Modern Languages at Cardiff University (Wales). Here we present the sequence implemented at the university.
The teaching sequence consisted of 30 teaching activities for students of Catalan as a foreign language at Cardiff University during the 2018-2019 academic year. The course lasted for two terms and the same group of students attended the classes of both terms (12 weeks, 36 hours of teaching each term). The first term corresponded to the A1 level and during this period both the PP and the imperfect were introduced. The second term corresponded to the A2 level, wherein the perfect was introduced and practiced in combination with the other two past tenses. Altogether, instruction of past tenses in the two terms took 9 hours in the A1 course and 3 hours in the A2.
A total of 10 students (aged between 19 and 29) in the A1-A2 levels participated in the sessions when the teaching sequence was implemented. Their L1s were English (7), Welsh (1), Portuguese (1), and Galician (1). All learners were studying other Romance languages at the university: they all had a high-intermediate (B2) level of Spanish, and some of them were also intermediate (B1) learners of Italian (1), Portuguese (3), and French (4). The main purpose of the implementation of the teaching sequence was not to examine the potential effect of the students' L1 but rather to design a sequence linking SLA findings and cognitive linguistics, and for this reason the issue of L1 influence is not discussed in this article.
The following sections focus on the design of the teaching sequence and the materials for the teaching of the meanings of the PP, the imperfect, and the perfect tense of Catalan. The forms were introduced in each case in class before the presentation of the meanings, but they were also practiced and reviewed during the sequence. The topics for all the language activities were determined by the syllabus of the course, following the course kits for the Module of Introduction to Catalan Language and Culture designed by the Hispanic Studies Department of Cardiff University. If not specified, all the activities and teaching materials were created by Ares Llop Naya.
The Catalan language classes (and the teaching sequence) followed a communicative approach, with an emphasis on both written and oral skills. At the end of each teaching unit, students had to prepare a written or oral final task integrating all the material practiced in the course.
Appendix 1 presents a detailed summary of the implemented teaching sequence, including the week (and the number of sessions) of teaching, the activity or materials used, and the tense-aspect instructional focus. In the following sections, the relevant aspects regarding the goal, design, and instructional strategies of the tense-aspect instruction are discussed in detail in the sequenced (chronological) order in which they were implemented in the classroom.

Introduction to the perfective past (PP)
The first verbal tense introduced to the students was the PP. It codifies the notions of past (in the metaphoric space of "there," far, and inaccessible) and finished (perfective) actions that contribute to the chronologic advance of episodes (Pérez Saldanya, 2000, p. 90). The main aim of the introductory activities for the PP was to secure the students' command of the meaning of this tense. One of the problematic aspects in the acquisition of this tense is that in Catalan, the periphrasis anar ('go') + infinitive is used to refer to past events. Two input flood and input processing activities were designed to introduce this tense. In Activity 1, students were presented the prototype of the PP (Figure 4). Activity 2 was an input-processing task with a two-choice picture-sentence matching task ( Figure 5) through which the prototypical meaning of the PP could be captured and contrasted to the prospective meaning. For instance, in the first sentence of Figure 5 (Va pujar la muntanya; 'He climbed the mountain'), students were supposed to select picture (b). The input forms were all telic predicates to reinforce early facilitative associations of perfective morphology and the lexical-aspectual properties of predicates (cf. the Aspect Hypothesis). Once the prototype for the PP was introduced, the focus of the teaching sequence was placed on the discursive use of PP as the verbal tense for events that constitute the foreground of stories (cf. the Discourse Hypothesis). This notion was presented to students in Activity 3 ( Figure 6).
The role of the PP as the tense to introduce the events corresponding to the foreground of narratives was presented to A1 level students through three activities. Activity 4 (Figure 7) was an input flood activity in which students had to put a list of sentences with PP forms in chronological order according to the plot of an edited cartoon video of an episode of Mister Bean they had to watch. This narrative was carefully selected following Blyth's recommendations (2005, p. 263), since the scenes included in the sequence were: (a) specific and easy to individuate (most of them included telic predicates); (b) prominent (as part of the foreground); (c) chronologically ordered; and (d) independent of the narrator's comments (without an evaluation of the events). The sentences that the learners had to organize chronologically following the video are presented here in the correct order (PP forms in bold; translations were not provided to the students): Figure 7. Activity 4: Picture-prompts based on a video for a picture-sentence matching input flood activity, introducing the prototype of the preterite

Introduction to the imperfective past (imperfect)
The second tense that was introduced to A1 students was the imperfect. It must be noted that acquisition of imperfect meanings is notoriously difficult for learners of languages that do not have an imperfective verbal marker (e.g., English or German). Furthermore, imperfective forms fulfil several functions that are not always easily comprehended by learners. In Catalan, imperfective forms have three basic meanings: habitual, progressive, and continuous (Pérez Saldanya, 2008). In the teaching sequence, since the focus was on narratives, we decided to begin with the continuous meaning because it is the most common one in the background of narratives. As for the case of the PP, after the introduction of the imperfect prototype in class (Figures 4 and 6), the first pedagogic strategy implemented was to flood learners with biased input that was coherent with the Aspect and Discourse hypotheses and the acquisition of prototypical combinations of morphology, lexical aspect, and discourse grounding. More specifically, the target bias was the association of imperfective atelic predicates with the background of scenes of narrative discourse.
Three sets of pictures were used as the main tool to illustrate the abstract conceptualization of narrative grounding. Each set reflected one of the different meanings of the imperfect tense (continuous, habitual, and progressive). For instance, Activity 11 ( Figure 8) was part of the set that was used to introduce the discursive component and presented the foreground and background notions. First, learners were asked what they could see in the picture. Their answers, as expected, focused on the actions taking place in the foreground, which in the plot of a narrative in the past tense would have been marked with PP forms. Next, students' attention was drawn to the background of the picture, using Figure 9, which was the same as Figure 8 except that the foreground and background events in the former were highlighted. More specifically, the foreground was marked in Figure 9 with temporally bounded events, whereas the background contained temporally unbounded states. The foreground was in color and contained a telic predicate marked with PP (La Jana i el gos es van amagar a la tenda 'Jana and the dog hid in the tent'), whereas the background was in black and white and included atelic predicates in the imperfect: Feia bon dia 'It was sunny'; Hi havia poc trànsit 'There was little traffic'; Les botigues eren obertes 'The shops were open'; and Els nens estaven contents 'The kids were happy'. Thus, the imperfect was presented in its continuous meaning as the form that referred to the identification of the state of things that constituted the background of a scene without clear temporal anchors. As mentioned in section 3.1., the purpose of these activities was to introduce the prototypical associations of the lexico-aspectual characteristics of predicates and discourse grounding (e.g., telicity and foreground) in simple sentences containing one verb to students who were exposed for the first time to the Catalan past tenses. As acknowledged by Blyth (2005), in the beginning stages of acquisition, the prototype needs to be introduced so that learners notice it and can process the new L2 meaning-form associations and ultimately produce them. In further stages, non-prototypical associations in complex sentences were introduced so that learners could become aware that the prototypical associations are not the only possible ones.
In Activity 12, students were presented with two sets of pictures (similar to Figures 8 and 9), in which the continuous and the progressive meaning of the imperfect could be elicited and shown to refer to the circumstances in which the foreground action took place. As most students were L1-English learners, their attention was drawn to the equivalent forms in English of the different meanings expressed by the imperfect in Catalan: past tense for the continuous meaning, used to + infinitive for the habitual, and was/were + V-ing for the progressive (Domínguez, Arche, & Myles, 2017, p. 434).
Activity 14 ( Figure 10) revolved around one of the key notions of the prototype of the imperfect: unboundedness. Unlike PP forms, which are completed processes and can thus be pinpointed in a timeline, the end of the event marked with the imperfect is "out of sight" and can refer to different time periods. To capture this notion, a multiple-choice activity was designed with a set of sentences with imperfective and present forms that could refer to any historical past moment (students had to choose if they referred to years 1900, 2000, or present). They were required to understand that, in theory, the imperfect could be used irrespective of how recent the events were. The contextual information in the sentence and the pictures allowed students to infer the correct answer. In Barcelona… a. 'There were few tourists.' b. 'The first cruises arrived.' c. 'People from Barcelona walked on horseback.' d. 'There is tourist massification.' e. 'The streets were not asphalted.' f. 'The citizens can travel by taxi.' g. 'The first shopping malls opened.' h. 'The citizens were happy for the celebration of the Olympic Games.' i. 'Family-owned business shops were passed down from generation to generation.' Activity 15 for A1 learners focused on the impact of imperfect forms on the meaning of unboundedness. As narrative grounding in scenes had already been explored, we focused on making students aware of what Castañeda Castro and Ortega (2001, p. 37) referred to as the "informative suspension" of the imperfect. The basic meaning of the imperfect in storytelling is mostly evaluative, and PP forms are needed for the story to progress. In Activity 15, learners had to read a list of sentences about the events in the background of the first cartoon video (Figure 7) and select what kind of information about the different scenes on the video could or could not be inferred according to the verbal tense used. For instance, in example (2), the form feia in the imperfect expressed the state of things in a non-specific, unbounded moment in the past; thus, the only information that could be inferred was the one in answer (b). In order to infer the option in (a) the verb in statement (2) should have been in the PP (va fer instead of feia).
'Someone opened the window to renew the air.' (b) No sabem si l'olor va desaparèixer o no.
'We don't know whether the smell vanished or not.'

Combinations of PP and imperfect forms at the A1 level
The third focus of the approach to teaching the Catalan past verbal tenses at the A1 level was the combination of PP and imperfect forms in narratives, a notoriously difficult task for beginner learners. The combination of verbal forms was presented using a metalinguistic awareness activity (Activity 18), in which students were required to reflect on the previous knowledge they had about the meaning of the two tenses. Learners were shown two pictures with a narrative plot containing visual marking of events in the foreground (Què va passar? 'What happened?') and states in the background (En quines circumstàncies? 'In what circumstances?') ( Figures 11 and 12). Students were asked to retrieve their previous knowledge about the PP and imperfect prototypes and their connection to discourse grounding to justify the use of the two tenses in the narratives. Most PP forms corresponded to telic predicates, whereas all the imperfect ones corresponded to atelic predicates. The exposure to this prototypical association of tense-predicate was in line, again, with the acquisition path of morphological forms according to the Aspect Hypothesis for the beginning stages of acquisition (see section 3.1.). Figure 11. Activity 18: Presentation of the use of PP to mark the foreground of a narrative containing PP and imperfect forms In the text of Figure 11, PP forms were in bold and encircled in blue (matching the images, in contrast to shadowed background images). Imperfect forms were just in bold. The English translation of the narrative is: 'He left the mountain hut early. It was sunny and the fields were green, and the trees had fruit. While he was walking, he saw three birds. He moved forward and found a horse, which was eating grass. While he kept walking, he came across a rabbit and saw how it was slowly walking away. When it was nighttime, he went back to the mountain hut again'. In the text of Figure 12 (same as Figure 11), imperfect forms were in bold and encircled in green, matching the images, while the PP forms were in bold.
Activity 20 ( Figure 13) was designed as a final and more complex activity in which learners were required to process and produce output. It was a dictogloss activity focusing on the students' awareness of the impact that the use of the preterite or the imperfect tense has in the overall meaning of sentences. First, students were given the visual information in Figure 13 and had to listen to the dictogloss story and take notes about it. Afterwards, in small groups, they had to rewrite the text they had heard paying attention to the verbal forms used. The text for the dictogloss was the following: Una parella es va casar. Estaven molt enamorats. Van anar de viatge a Mallorca. Feia molt bon temps. Durant el viatge es van enfadar. Ell estava molt enfadat i, per això, va tirar l'anell de casat al mar. Quan van arribar a la ciutat estaven molt enamorats de nou i volien fer una festa. Van anar al mercat. Al mercat hi havia molta gent, però van comprar un bon peix per cuinar-lo. I quan van tallar i obrir el peix ... Sabeu què hi van trobar? Les espines! [English translation (not provided to the students): 'A couple got married. They were very much in love. They went on a trip to Mallorca. The weather was very good. During the trip they got angry. He was very angry and, because of that, he threw his wedding ring into the sea. When they arrived in the city they were in love again and they wanted to celebrate a party. They went to the market. There were a lot of people in the market, but they bought a fish to cook it. And when they cut and opened the fish … Do you know what they found inside it? The fishbones.'] Figure 13. Activity 20: Visual support material for students to reconstruct the Dictogloss story In Activity 20, students were encouraged to debate and discuss in small groups the grammatical and pragmatic acceptability of certain combinations of the dictogloss text, using garden-path sentences; that is, sentences that are grammatically correct even if, at first glance, might be misunderstood or perceived as incorrect (Alonso-Aparicio, 2014, p. 13, following Doughty & Williams, 1998).
Finally, the goal for Activity 21 was to show students that even if temporal markers can contribute to a better understanding of a specific verbal tense, the link between verbal tenses and temporal markers is not exclusive (Llopis-García et al., 2012;Alonso-Aparicio & Llopis-García, 2019). Through a structured output activity, learners became aware that the meaning of a sentence can change depending on the use of a verbal tense together with a temporal marker. Learners were given pairs of sentences with a blank that they had to fill in with a PP or an imperfect form and then associate to an interpretation. For instance, in example (3), students were supposed to use the verb anar, 'to go' in the PP (vam anar, 'went') for interpretation (a), and in the imperfect (anàvem, 'used to go') for interpretation (b), even though both sentences included the same time expression (cada any, 'every year').
(3) To wrap up all the concepts learned in the A1 course, students did an output final task, in which they had to create a written narrative timeline with the most representative events and circumstances of the life of Catalan musician Pau Casals. Students had previously read an adapted biography in Catalan (Bayà, 2004) and were familiar with Casals' life and related vocabulary. Feedback to students' written productions was given individually and commented on with the whole group during the first session of the second semester (beginning of the A2 level course, Activity 23).

Teaching sequence for the instruction of past tense-aspect in L2 Catalan: A2 Level
Students in the A2 level course at Cardiff University were the ones who had followed the first part of the instructional treatment during the first half of the course, and they had a good command of the concept of perfective-imperfective aspect and the idea of the external observer's perspective with respect to finished actions. As was done for perfective and imperfective aspect, the first activity of the A2 level course consisted of the presentation of the prototype for the perfect tense (Activity 24, Figure 14). The key notion that was highlighted was that, in the case of the perfect tense, the observer's perspective is internal (aquí, ara, espai actual; 'here', 'now', 'actual space'), because the finished action is relevant to the speech situation. In this respect, following Castañeda Castro (2014, pp. 279-280), we aimed to evidence the flexibility of the notion of relevance with respect to the actual time: a perfect tense can refer to a current situation, to a part of the current day, month, or even to someone's (alive) whole life. This general overview of the construal of the perfect was useful to explain by means of a learner-accessible rule the experiential, resultative, and inclusive meanings of the perfect in Catalan (Pérez Saldanya, 2008). In this respect, Catalan is not different from the other Romance languages. However, the Catalan perfect has a strict hodiernal use, that is, it is used for all actions that have happened within the day of the speech situation, and such use is uniform, in contrast with other Romance languages (see Section 2).
A total of seven activities for the introduction and practice of the perfect were designed. In this section, we discuss six of them. The first meaning of the perfect that was presented to the students was the experiential one, because this meaning is homogeneous in other Romance languages and exemplifies in a clear way the notion of past connected with the present. In Activity 25, A2 students divided into groups of three were given a card with a list of questions about experiences or actions. Each of the cards (e.g., Figure 15) referred to a different time period that was connected to the present time ('your life', 'this week', and 'today'; avui, 'today' in Figure 15). This temporal contextualization was included as a title for the timeline to contextualize each card, but not in the sentences or answers given as input (Figure 15). The first part of activity 25 consisted of an input flood activity (asking a question including a perfect form) combined with a structured output production (answering the same question using the same perfect form). At the end of the activity, students were asked to reflect on the meaning conveyed by perfect forms. They commented on why, irrespective of the timeline extension (i.e., your life, this week, or today), the tense used in all the cards of the activity was the perfect.
Activity 26 consisted of a complementary two-phase activity. It was created to assist A2 learners to capture subtler instances of the combination of two meanings of the perfect: (a) current relevance and (b) finished action (i.e., perfective). The first part of the activity (26a) consisted of the matching of pictures of a minimal pair with the meanings of specific sentences containing a perfect form ( Figure 16). For instance, for the minimal pair in 1 in Figure  16, the appropriate picture depicting Ha dinat, 'He has had lunch' (perfective meaning of the perfect, in a hodiernal use) was Picture (a). For the minimal pair in 4, the appropriate picture for sentence Ha estat a París, 'He has been to Paris' (experiential perfect) is Picture (a). Note that Activity 26a included the hodiernal, resultative, and experiential meanings, but not the inclusive, which was practiced in other activities. Some of the pictures admitted two possible answers. For instance, for picture 4, the use of the perfect could be hodiernal (Today he has been to Paris) or experiential (He has been to Paris). Both answers were discussed in class and considered appropriate. Figure 16. Material for input processing activity about the meanings of the perfect tense Translations: 1. 'He has had lunch'; 2. 'It has rained'; 3. 'He has fallen down the stairs'; 4. 'He has been to Paris'.
After Activity 26a, the correct comprehension of the meanings of the perfect was reinforced using three adapted pictures of the prototype illustrating each of the meanings (Figure 17, resultative meaning; Figure 18, experiential meaning; Figure 19, inclusive meaning). During the second part of Activity 26 (26b), graphic representations of perfect forms were substituted for sentences referring to discourse contexts. Such contexts did not include any perfect forms, and students had to match them to other sentences containing a perfect form according to the information conveyed or potentially inferred from the contextual clues. For instance, students were given sentences such as the ones in (4): (4) 1. No pot sortir de casa.
'He can't leave home.' 2. Busco una cambrera amb experiència. 'I am looking for an experienced waitress.' And they had to match them to provide sentences containing perfect forms such as the ones in (5): (5) (a) Ha nevat molt.
'It has snowed a lot.' (b) Ella ha treballat a cinc restaurants diferents. 'She has worked at five different restaurants.' After the general overview of the different meanings of the perfect was completed, the next section of the instructional procedure was centered on the hodiernal (perfective) use of the perfect in Catalan. The hodiernal use in Catalan was reinforced in Activity 27, an input flood activity, followed by an input processing activity. Students were shown an authentic video from a Catalan Tourist Board campaign (Catalan Tourism Agency, 2018). The video did not include any text and consisted of a sequence of scenes about the activities a couple of friends had done in Catalonia to return a lost travel journal to its owner ( Figure 20). Students were given a list of sentences with perfect forms and had to identify the true and false sentences, according to the images shown in the video. As the target meaning was the hodiernal, the instructions given to students implied that the situations in the video had taken place during a whole day (corresponding to avui, "today"). Some of the sentences that were provided to the students were: 1 After commenting on the true and false answers, students were asked which specific day (and why) they thought the story had taken place, according to the sentences given. The target answer was "today," and the use of the present perfect was thus justified.
The sequence of the episodes shown in the video for Activity 27 was used as a precursor to connect the meaning of the perfect with its discursive dimension. The key information we expected learners to capture was that the perfect, as a perfective tense, is associated with the foreground of chronological scenes. The instructional treatment aimed to make students aware of the underlying conceptual differences between the PP and the perfect. Thus, Activity 28 consisted of a reading comprehension activity of an extract of a personal diary (from published materials: Vilagrasa, 2018), in which the contrast between the PP and the perfect was transparent; that is, PP for pre-hodiernal events and perfect for hodiernal events: (6) Fa una setmana que vaig firmar el contracte de lloguer del pis I que em van donar les claus … Va ser fantastic! Avui, en canvi, ha sigut un dia de contradiccions… 'A week ago, I signed the rent lease for the apartment, and I was given the keys … It was wonderful! Today, on the other hand, has been a day full of contradictions…' The grammar activity (28a) consisted of justifying the use of each past form in the diary text and explaining the different meanings they conveyed. Once the distinction was clear, in the second part of the activity (28b), students were asked to write some additional extracts to elaborate on the background of the story of the text. The target tense for this part of the activity was the imperfect and paralleled the use of PP and imperfect in Activity 20 but this time using perfect (hodiernal use) and imperfect for a hodiernal narrative within a personal diary.
After additional practice of verbal forms (Activity 29), A2 students were presented with the individual final task of the sequence. Activity 30 was a free output production activity due at the end of the fourth week of the second semester. It was designed to let students practice the three past tenses in addition to other components of the course syllabus corresponding to the unit about the description of objects and works of art (week 3). Each student was asked to write a short history of an object or a piece of clothing that was special for them. They were asked not to include their name in the written text, so that the rest of the class could guess whose element was described. The instructions given to students included specific questions to elicit the use of each of the three past tenses (PP in Question 1, imperfect in Question 2, and PP and perfect in Question 3; instructions were only in Catalan; here, we provide the English translation (7)): (7) Explain the history of a special object or piece of clothing you had when you were a child, or you were younger, without telling your identity. Try not to provide obvious tips on who you are. Comment on the following: This final task in the A2 course concluded the implementation of the teaching sequence for the instruction of the three past tense forms and meanings in L2 Catalan.
The goal of the design and implementation of the teaching sequence was not to conduct an experimental study, and thus, evidence of acquisition was only gathered in an informal way. Three types of evidence point to a positive attitude from students regarding the implementation of the teaching sequence: comments from the students, written productions for the final task at the end of the A1 course (Activity 22), and written productions for the final task at the end of A2 (Activity 30). In general, students received positively the cognitive approach and several of them mentioned to the teacher that the prototypes were helpful to understand the general notions about tense-aspect in Catalan and how they contrasted with other languages: (8) Els dibuixos van ser molt útils perquè explicaven molt clarament quan s'utilitzen els temps verbals en català, perquè són una mica diferent als del castellà. The pictures were very useful for me because they clearly showed when the verbal tenses had to be used in Catalan, as they are a little bit different from the ones in Spanish.' (9) Recordo que tu ho vas explicar molt i molt clarament. I hi havia un dibuix de un poble/parc que vas utilitzar per explicar quan s'utilitzen els temps diferents; per exemple amb activitats o sentiments ... i per capes del discurs.
'I remember you explained that very very clearly. There was a picture of a village/park you used to explain when the different tenses are used; for example, with activities or feelings … and according to discourse grounding.' Qualitative evidence from the final tasks in the sequence also showed that students understood well the use of past morphology. Some excerpts from the timeline biography of Pau Casals (Activity 22) show how the A1 learners, who had been exposed to 36 hours of Catalan instruction, could combine the use of PP and imperfect in an accurate manner, even though there were some instances of overuse of the imperfect instead of the PP (marked with *): (10) Als noranta anys, Casals va reprendre la vida pública. Va fer gires per tot el món. L'any 1971 va rebre la medalla de la pau de les Nacions Unides, on va fer un discurs molt important davant tots els dirigents polítics. At the end of the final task in A2, learners produced a written text in which they told the history of an object (Activity 30). The two following excerpts from two of the texts are instances of the type of texts they produced (Excerpt 13, with 6 accurate forms: 2 PP, 3 imperfect, and 1 perfect; and 1 PP inaccurate form; Excerpt 14, with 5 accurate forms: 1 PP, 2 imperfect, and 2 perfect; and 1 inaccurate PP form): (12) Per a mi, el casc era molt important ja que em va permetre muntar una moto per primera vegada, i he muntat motos des d'aleshores. El casc significa molt ja que recupera records feliços de quan jo era jove i el meu pare i jo *va anem a una pista local de terra. M'encantava portar el meu casc perquè pensava que era tan bonic! Una altra raó per la qual el casc és tan especial és que va inspirar el meu amor actual de les motos i les carreres. 'For me, the helmet was very important because it gave me the opportunity to ride a motorbike for the first time, and I have ridden motorbikes since then. The helmet is so important, because it brings back fond memories from when I was a child and my father and I went to a local track. I loved wearing my helmed because I thought it was so beautiful. Another reason why the helmet is so important to me is that it inspired my current passion for motorbikes and races.' (13) He elegit la meva manta com l'objecte especial que descriuré.
[…] em recorda molts moments especials i té valor sentimental perquè la meva àvia ho va fer per a mi. Ho he tengut des que tenia tres anys i sempre la portava tot arreu. La meva mare solia ocultar-la de mi perquè *va tenir que rentar-la. 'I have chosen my blanket as the special object I am going to describe. It brings back special memories and has a sentimental value because my grandmother made it for me. I have had it since I was three and I always used to bring it everywhere with me. My mother used to hide it from me because she had to wash it.'

Discussion
The primary objective of this article was to design a teaching sequence informed by SLA findings and CL. Thus, we adopted Ellis's (2005) recommendation of examining teaching problems in light of SLA. In this section, we discuss some of the advantages and drawbacks of this type of teaching. An advantage of examining a teaching problem from an SLA perspective is that if the teaching problem, in this case the acquisition of tense-aspect, has been extensively researched, the potential application of some of the findings is not as difficult as in cases with little research. The L2 acquisition of tense-aspect is well researched and thus the integration of such findings into the sequence was not a complex matter. The design of the teaching sequence shows that SLA findings were useful in making decisions on many issues, such as the order in which to teach the tenses (PP, imperfect, combination of PP and imperfect, perfect, combination of all tenses), the role of providing quality input before producing output, the intervening roles of lexical aspect and discourse grounding, as well as the effort to provide operational rules (i.e. prototypes presented via drawings; see  instead of lists of rules and associations of temporal markers with specific verbal forms. Having access to SLA knowledge also proved fruitful in the sense that it gave a sense of security for the teacher regarding what to expect about the interlanguage development of learners and how to approach decisions regarding the order of teaching and justifications for tense use. Regarding the application of CL to teaching tense-aspect, as reviewed in section 1.2, proponents of CL have argued for two main advantages: a cognitive approach is closer to the learners' cognition and thus might facilitate language acquisition, and it helps avoid providing long lists of specific meanings and associations of adverbials with tense-aspect meanings. However, as evident in the diverging opinions in the quotes in section 1.2, the responses toward the applicability of CL range from full enthusiasm to skepticism. In addition, empirical studies on the comparison between traditional and cognitive approaches have provided mixed results. On the one hand, such results can be considered positive because in all studies the cognitive treatment group outperformed the control group, which implies that the participants learned what was taught about tense-aspect. On the other hand, the differences between the traditional and cognitive groups tended not to be significant. The reasons for this have been related to several issues, such as the complexity (in terminology and notions) of the cognitive approach, the impossibility of the prototype framework to account for important exceptions and non-prototypical matchings between stative and activity verbs with perfective markings (Rastelli, 2020), the level of learners, the novelty of the approach, and motivational issues (i.e., lack of interest in the cognitive approach; Bielak & Pawlak, 2013;Kermer, 2016).
Considering the previous observations, in order to advance in the teaching of L2 tense-aspect that is informed by SLA and CL, further studies need to concentrate at least on four main issues. First, research should focus on how the individual characteristics of learners and teachers are affected by the classroom implementation of units based on CL. While the learners in Bielak and Pawlak's (2013) study did not seem very motivated by the implementation of CL, the learners who were exposed to the sequence in the current article seemed to have a positive attitude toward it. This raises the issue whether applied cognitive grammar is suitable for all types of learners. Previous studies have argued that a cognitive approach might be more adequate for higher level learners (Bielak & Pawlak, 2013;Kermer, 2016), but the current study suggests that it can be used with beginner learners as well. It could be argued that the learners at Cardiff University were university learners who already knew other Romance languages and thus their previous L2 knowledge had a positive transfer effect in their acquisition of Catalan. While this might have been an influence, a similar sequence was implemented with a multilingual class at the Barcelona Official Language School with students of diverse backgrounds and L1s and the reactions were similar.
Second, further research in the application of CL in language teaching needs to investigate the effects of different designs of teaching sequences and how they are implemented in the classroom. Further quantitative and qualitative studies need to investigate different types of sequences of activities as well as the optimal number of activities and the method of implementing them in different types of classroom teaching (e.g., communicative, task-based, or grammar-based). The current implementation showed that it is possible to embed cognitive grammar explanations within a communicative approach that incorporates information about discourse grounding and verbal semantics. The communicative aspect of the course was clear in the design of the two final tasks (biography of Pau Casals and history and description of a symbolic object) and the integration of all skills. In sum, adopting a CL approach does not mean turning the L2 classroom into a grammar session, or at least it does not turn it any more than any other approaches (Niemeier, 2017).
Third, what characterizes the sequence that has been presented is that it includes numerous activities compared to the low number used in some studies on the use of CL and language learning (30 activities within a period of 9 hours for the A1 students in this article, compared to 3 sessions for a total of about 3 hours and a half in Alonso & Llopis-García, 2000; 1 session of about 90 minutes in Bielak & Pawlak, 2013; 45 minutes per week in 3 consecutive sessions in Kermer, 2016). The number of activities and time devoted to teaching in the studies is not comparable, because the current study was not experimental, and the time in the classroom was also devoted to other matters, e.g., discussion of vocabulary (see Table A1). However, further research could investigate the extent to which the acquisition of complex systems such as the tense-aspect one requires high exposure and practice. For instance, special attention needs to be paid to the number of tenses that are taught (three, in the case of the current sequence) and how they are sequenced but presented as a coherent system grounded in the notions of space of the action, the action of the verb, and the observer's perspective.
Finally, further research needs to investigate whether the findings of previous research regarding better results for long-term effects of cognitive grammar compared to better results for traditional methods in the short run are confirmed or not (Bielak & Pawlak, 2013, p. 218).
Regarding the drawbacks of the adoption of CL for language teaching, applied cognitive grammar can be a double sword for language teachers. On the one hand, a CL account of language might be beneficial because it can motivate and explain the use of tense-aspect systems in a unified manner. On the other hand, the complexity (in terminology and in adopting new conceptualizations) can be too high for some language teachers. Tyler (2012, pp. 221-222), in a section aptly titled "Putting linguistics back into L2 research and teaching" summarized the conundrum effectively and reached the conclusion that the effort is worthwhile: L2 professionals need as complete an understanding of grammar and lexis, and the motivated ways in which they are used within communication, as possible. In order to achieve and use this understanding, they are best served by a model of language that is as accurate, accessible and as complete as possible.
[…] CL is an approach to language that meets these criteria, and hence that the conceptual insights and analytical tools it provides have a good deal to offer L2 researchers, L2 teachers and, above all, L2 learners. […] L2 professionals have good reason to be wary of embracing an unknown model of language, particularly given the time and effort that will be necessary for them to gain a deep enough understanding of the approach to apply it to their research and teaching. Thus, the overarching aim […] has been to provide evidence that CL is an approach to language that is 'worth the pain.' For the effort to be even worthier, future research in language teaching pedagogy needs to find ways to adapt CL principles into language teaching materials that are appropriate for learner characteristics, such as their level of proficiency and motivation. In this respect, pedagogical grammars that adopt a cognitive or cognitive-inspired account (e.g., Alonso Raya, Castañeda Castro, Martínez Gila, Miquel, Ortega, & Ruiz Campillo, 2011;Bastons, Bernadó & Comajoan-Colomé, 2011;Petri, Laneri, & Bernardoni, 2016) or recent works that examine SLA and pedagogy within a cognitive perspective (Ibarretxe-Antuñano, Cadierno & Castañeda Castro, 2019;Masuda, 2018;Niemeier, 2017) are a welcome and necessary addition.

Conclusion
In this article, we adopted the perspective of examining a teaching challenge (L2 tense-aspect) for most L2 teachers from the lens of SLA research and CL. The reviewed studies on cognitive grammar and applied cognitive linguistics point to a situation in which cognitive grammar is in the position of bringing in a new perspective of the teaching of tense-aspect but that has not been properly empirically validated yet. In this respect, further research needs to provide evidence for the cognitive approach keeping in mind that adopting a cognitive approach might not be easy neither for teachers nor learners due to the change of perspective, types of activities, and terminology. For cognitive approaches to be relevant, teachers need to be exposed to them, more research is needed, and more pedagogic experiences need to be designed. In this respect, we have provided a teaching sequence for the teaching of Catalan that can be adapted for other Romance languages that have similar tense-aspect features. We hope that the detailed description of the design and implementation of the teaching sequence that combines SLA findings and CL will inspire language teachers to create new connections between theory and practice and will contribute to more effective language teaching.

A2 Level
Week Session (hours) Activity/materials Tense-aspect instruction focus 2 1 (2 hours) 23. Group discussion about the written final tasks.
24. Presentation of the prototype of the perfect. 25. Input flood, input processing and structured-ouput production. Question-answer cards.
26a. Input processing. Picture-sentence matching activity. 26b. Input processing. Sentence-discourse matching activity. 27. Input flood and input processing activity about the hodiernal use (audio-visual source) combined with a true-false activity.
23. Feedback about the final A1-level task and comments on the use of the two past verbal tenses studied in the A1 course. 24. Meaning of the prototype of the perfect. 25. Focus on the importance of the notion of current relevance of finished actions to understand the prototype of perfect forms. 26a and 26b. Indirect exposure to the resultative/experiential/inclusive meanings of the perfect tense. 27. Hodiernal use of the perfect in Catalan.
2 2 (1 hour) 28a. Input processing. Reading comprehension activity to focus on the contrast between the perfect and the PP, and comparison of the prototypes of the PP and the perfect. 28b. Output processing. Short writing activity to add a background (in the imperfect tense) to the story provided in 28a. 29. Focus on meaning and form activities as general review.
28a. Use of the PP and the perfect as the tenses for the events in the foreground of scenes, contrast of the PP and perfect prototypes.
28b. Reinforcement of the distinction of foreground and background and its association with past tenses.
29. Review of all tenses 4 Final A2 task 30. Final task. Output production. Narrative about the history of a special object.
30. Practice of the three tenses.