TEACHERS’ CONCEPTS ON THE TEACHING OF GRAMMAR IN RELATION TO THE TEACHING OF WRITING IN SPAIN: A CASE STUDY ANNA CAMPS*, XAVIER FONTICH* *Universitat Autònoma de Barcelona Abstract This article presents the results of a study conducted in Spain concerning a secondary school teacher’s conceptualizations of grammar teaching and its relation to writing. Concepts are considered of the utmost importance when confronting any educational change. In dealing with the controversial issue of the role played by grammar in learning to write, the authors defend their positioning of the debate around the concept of metalinguistic activity as a source of grammar learning rather than the notion of an implicit/ex- plicit grammar dichotomy. A number of studies have deemed the former as a promising avenue for re- search. After analysing a semi-structured interview held with the teacher, some preliminary results con- cerning the respondent’s concepts show that she combines ideas of implicit grammar and explicit gram- mar with an incipient awareness of the importance and possibilities of promoting metalinguistic activity. Awareness of metalinguistic activity can anchor the re-orienting of classroom practices towards reflection within planned activities designed by the teacher and intended to promote students’ autonomy and awareness. Nonetheless, these results point at a much-needed process of structuring by means of which teachers enhance this perspective. Keywords: grammar teaching, writing, implicit/explicit grammar, metalinguistic activity, teacher’s con- cepts 1 Camps, A. & Fontich, X. (2019). Teachers’ concepts on the teaching of grammar in relation to teaching of writing in Spain: A case study. Contribution to a special issue What is Grammar in L1 Education Today?, edited by Kaisu Rättyä, Elżbieta Awramiuk, and Xavier Fontich. L1-Edu- cational Studies in Language and Literature, 19, 1-36. https://doi.org/10.17239/L1ESLL- 2019.19.02.02 Corresponding author: Xavier Fontich, Universitat Autònoma de Barcelona, Departament de Didàctica de la Llengua i la Literatura, i de les Ciències Socials, Facultat de Ciències de l’Educació, Edifici G5, Campus UAB, 08193 – Bellaterra (Cerdanyola del Vallès, Barcelona), email: xavier.fontich@uab.cat © 2019 International Association for Research in L1-Education. 2 A. CAMPS & X. FONTICH 1. INTRODUCTION A frequent statement in current language curricula in some countries is that the grammar content that is to be taught must be related to the use of the written lan- guage in order to improve it. Yet, some far-reaching research on the possible inci- dence of school grammatical knowledge in the progress of students’ competence in writing concludes that this purpose has little empirical evidence and that, regardless of content and methodology, grammatical knowledge does not contribute to (or scarcely influences) the improvement of written uses (see Andrews, 2010; Graham and Perin, 2007; Hudson, 2001; Wyse, 2001). The exception for this is “sentence- combining”, which has proved to be an effective procedure for improving writing (see Andrews et al., 2004; Graham and Perin, 2007). While a number of studies main- tain that the writing-grammar interrelation is affirmed but not verified, a research trend during the past decade has contributed empirical evidence about this connec- tion and its difficulties, along with theoretical revisions as to how to best approach it (Dolz & Simard, 2009; Fontich & Camps, 2014; Locke, 2009 & 2010; Myhill et al., 2012; Myhill & Jones, 2015; Myhill et al., 2016; Rodríguez-Gonzalo, 2015; Van Rijt and Coppen, 2017). Nonetheless, the grammar-writing issue remains controversial. In Spain, various attempts to reform grammar, far from helping to bridge the gap, have simply dis- tanced the study of grammar still further from writing (see Fontich & García-Folgado, 2018, and González-Nieto, 2001 for a revision). The contribution of linguistic struc- turalism and generative grammar consisted, above all, in offering schools a greater conceptual precision in the definition of grammatical categories and systems of anal- ysis, but this reified the isolation of grammar content with regard to language use, culture, and writers’ intentions. In order to overcome this, school instruction re- sorted to linguistic orientations aimed at the study of text and discourse; thus, con- cepts that could account for the relationships of the formal elements beyond the sentence were introduced, namely text-level concepts (e.g., anaphora, correlation of the verb tenses, etc.) and enunciative-level concepts (e.g., modality, etc.) (Gonzá- lez-Nieto, 2001)1. While this represented a shift in how to relate adequate linguistic content to teaching, two aspects remained unresolved: how to integrate morpho- syntax as an instrument for reflection on texts and discourse, and how to re-think not only content but also methodology (Fontich & García-Folgado, 2018). Therefore, a major question in Spain is “What are the obstacles that stand be- tween learning and improving the formal uses of the language and grammatical 1 In Spain, for example, according to the so-called LOMCE for linguistic education in primary and secondary schools (Organic Law 8/2013 for 9 December 2013), the teaching of grammar needed to “move away from the attempt to use linguistic knowledge as a goal in itself to re- store its original functionality: to serve as a basis for the correct use of the language” (p. 1). According to this text, linguistic competences are learned “from the knowledge and reflection necessary to appropriate grammatical and orthographic rules, essential to speak, read, and write correctly in all spheres of life” (p. 1). TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 3 knowledge?” Some lines of research of a qualitative nature have enabled the point- ing out of ways towards a possible answer. On the one hand, research on students’ grammatical concepts demonstrates the inconsistency and lack of systematization of the concepts that students elaborate (e.g., Casas, 2014; and see for similar results Myhill, 2000 in the UK, and Fisher, 2004 and Nadeau & Fisher, 2011 in Francophone Canada). This lack of systematicity would be closely related to the formulations of textbooks that teachers and students use as guides to learn grammar, and, for in- stance, would manifest itself in tasks of identification (Coronas, 2014). Another im- portant problem regarding grammatical concepts is the fact that there is limited agreement among linguists about the relevant concepts that should be taught or used to enrich traditional grammar education (see Van Rijt & Coppen, 2017). On the other hand, however, some of these investigations (many of them based on interview-about-instances to students) have shown the students’ remarkable abilities to take the language as the object of their observations and their comments (i.e., their ability to carry out a metalinguistic activity; see section 2 below and Fon- tich, 2014). (For how student teachers deal with conceptual systems see Gil, 2017 for the Spanish context, and Nupponen et al., in this special issue, for the Finnish context). Likewise, another line of research, based on the in-depth study of instructional sequences on writing and grammar, also allows for verifying the possibilities offered by reflexive writing and reflection on grammatical issues for unleashing the metalin- guistic activity of students (see Milian, 2005; Rodríguez-Gonzalo, 2015). These stud- ies maintain above all that grammar learning is the result not of direct instruction but of a teacher’s mediation and a framework sustained overtime. Rather than ap- proaching the teaching process as a set of ready-made strategies, the teacher pre- pares a structure that can be adjusted to the students’ idiosyncrasies, difficulties, and strengths (see Engeström, 2011). This set of investigations suggests that locating the problem in the field of met- alinguistic activity represents an important shift in the conceptualization of how we might approach the problem at issue (Fontich & Camps, 2014). Metalinguistic activity would be any verbal or non-verbal activity that focuses on language itself as its object and that allows for observing and studying the process of grammar learning (Ribas et al., 2014). It is the kind of reflective practice carried out when we review a text, correct our own or a speaker’s verbal expression, identify elements within a sen- tence, or adapt our text to the addressee or to the requirements of the genre, etc. (Fontich, 2016; Gombert, 1990; Taylor, 2000; see also Van Rijt et al., in this special issue). With respect to writing, according to Myhill & Jones (2015) it is “always an act of selecting, shaping, reflecting, and revising […] and thus draws on metalinguistic activity” (p.840). Within such a perspective, the link between metalinguistic activity and grammar teaching leads to the following assumption: as in all fields of knowledge, the learning of concepts (in our case, word-, sentence-, text-, and discourse-level grammar con- cepts) can offer students tools to facilitate such reflection. That is, rather than a 4 A. CAMPS & X. FONTICH direct and immediate transfer of concepts into practical resources, the activity of consciously using the language will involve processes of reflection about such lan- guage use, as well as elaboration of concepts through a complex process of concep- tualization: from use to abstraction and from abstraction to linguistic use, via the teacher’s mediation (Rodríguez-Gonzalo, 2015). In other words, while this two-way process based on metalinguistic activity can be carried out spontaneously by the stu- dents on their own, it also allows for being scaffolded with the teacher’s assistance –especially by way of guided induction (see Simard et al., 2010; and Gil & Bigas, 2014). In this respect, teachers’ conceptualizations have been regarded, in turn, as core aspects that guide pedagogic action and as a paramount issue to be dealt with in research on teacher education (Watson, 2015a and 2015b). Here is where we locate the study presented in this paper, with some results from our analysis of a teacher’s concepts concerning the teaching of writing and grammar (for other results within the same research project, see also Fontich & Birello, 2015; Fontich & Camps, 2015; and Bastons et al., 2017) 2. Section 3 describes the basic methodological layout of the study and section 4 some of the general results are briefly presented, enabling us to extensively focus on the analysis of one of the in- terviews (section 5). First, in section 2, we address some basic notions of the con- cepts of explicit/implicit grammar knowledge and of metalinguistic activity that will serve to frame the analysis. 2. METALINGUISTIC ACTIVITY, A TWO-WAY STREET BETWEEN WRITING AND GRAMMATICAL KNOWLEDGE In Spain there has been a general tendency to raise proposals for reform of the teaching of grammar from a reconsideration of the content. This can be seen in the generalization of the structuralist and generativist models of the 1970s, models that advocated a rationalization of scholastic grammar concepts. The latter were based on the so-called traditional grammar principle, not deemed as particularly rigorous since it mixed semantic, pragmatic and formal criteria when dealing with grammar description (Cuenca, 1992; Tuson, 1980). Although from a very different approach, new linguistic-discursive approximations, which were extended to the grammar of the text and discourse, also represented a shift in the content proposed for teaching (González-Nieto, 2001). In this case, the shift was based on an approach focused on the need to promote the improvement of students’ texts. This led first to focusing on the mastery of different text types, and subsequently, to addressing the diversity of discursive genres (e.g., Abad 2015; Castellà, 1994; Hernández-Navarro & Castelló, 2 The study is part of the funded research EDU2011-26039, “La incidencia de la reflexión gramatical sobre la lengua en la construcción de la competencia escrita” [The incidence of grammatical reflection on the language in the construction of written competence], with Dr. Teresa Ribas i Seix as its PI and financed by the Spanish MINECO (Ministry of Economy, Industry and Competitiveness). TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 5 2014), a trend also developed in other countries and regions (e.g., the Hallidayan tradition in Australia; see Jones & Derewianka, 2016). However, a general lack of improvement in students’ texts led to thoughts that both content and teaching pro- cedures ought to be revised (Camps & Ferrer, 2000). This directed many teachers and researchers towards expanding the need to review the teaching of grammar, by considering not only what the most adequate grammar contents might be but also taking into account the teaching and learning processes involved (i.e., methodologi- cal practices, learning obstacles). This section is oriented towards offering a framework to locate the teaching of grammar within the interplay between content and teaching-learning practices―the so-called pedagogic system (see Fontich & Camps, 2014). In the first place, it consid- ers what is meant by grammar, which will lead us to reflect on two of its meanings (namely, implicit and explicit grammar). Discussion of the idea that learning grammar in the framework of L1 consists of making explicit the grammar that we know implic- itly, concludes this first part. Secondly, the concept of metalinguistic activity that is considered the source of grammatical learning is addressed both in the development of verbal uses and in school learning of grammatical concepts. Finally, a model of grammatical teaching and learning, based on reflection and metalinguistic activity, will be advocated. 2.1 Implicit grammar and explicit grammar In the field of teaching, two concepts of grammar have been generalized, and also underlie the most frequent definitions of general dictionaries and handbooks of lin- guistics (Fontich & Camps, 2014; Tuson, 1980). The first concept of grammar is that of implicit grammar, understood as the sys- tem of form combinatory that the speaker commands and that allows the producing and understanding of a certain language (i.e., learning to speak involves learning the verbal elements and their combinatory in a largely unconscious way). It can be said that the child learns (some authors would say that she acquires) the grammar of her language (cf. Lantolf & Poehner, 2014). So do adults when they learn a second lan- guage or additional languages in a natural situation, although, as we see later, it would be necessary to clarify what it means to learn unconsciously. The interpreta- tion of the origin of this internalized grammar is different according to linguistic schools. On the one hand, some approaches adopting a cognitive perspective (e.g., generative studies) consider it to have an innate component, which according to Bronckart (2008) requires that “the processes of noesis (of pure thought) are primary with regard to the processes of semiosis and independent and autonomous with re- spect to the latter” (p. 7, italics in original)3. This natural foundation underpins the 3 Even though mainstream generative grammar innateness (in the form of a separate lan- guage faculty) is not shared by all the contributions within the large family of generativist 6 A. CAMPS & X. FONTICH stability of units of thought and the actual languages would be assumed not to play any part in their formation (i.e., the actual languages would be epiphenomenal). Con- trariwise, some other approaches (e.g., rooted in a sociocultural framework) con- sider that grammar is, on the other hand, constructed in a process of progressive structuring of the input that learners receive from the environment. It could be said that while for the former perspective grammar allows for communication, the latter entails that it is communication that triggers a sedimented set of rules that we call grammar knowledge―that is, while rules may help to explain how a language works, rules are not the origin of languages (see Hopper, 1998; and see also Fontich, 2016 for a discussion). The second concept of grammar falls into a different perspective. Languages have been the object of study and theorization, and such study and theoretical systema- tization of the structure of languages is also called grammar (Bernárdez, 2008 and 2017; Harris & Taylor, 1989). This is the case of explicit grammar conceived as the science that studies and describes the structures of a certain language (e.g., French grammar, Spanish grammar), as well as, more restrictively, the theory or model of analysis of the structure of languages (e.g., structural, generative, cognitive... as well as pedagogical grammars; Cuenca, 1992). Thus, it is applied to the elaborated and systematized knowledge about the language, its structure and its functioning. Ac- cording to Taylor (1997), the possibility of such systematization is afforded by lan- guages themselves via reflexive meta-discourse for which “Language could be said to have, not an internal, but an external skeleton” (p. 12). This grammar is conscious and it implies conceptualization and systematization as well as the use of specific terms to name the concepts. Furthermore, it is not learned spontaneously or natu- rally, but must be taught. 2.2 The implicit / explicit grammar dichotomy in the teaching of languages The dichotomy between both grammar concepts has had an enormous impact on the teaching of languages and has been applied both to the knowledge involved in what is meant by knowing grammar, and to the learning and teaching of both second and foreign languages as well as of a first language. In this paper we focus on the latter4. models (e.g., J. Bresnan and R. Kaplan’s Lexical Functional Grammar LFG; see Nordlinger & Bresnan, 2011). 4 Overcoming a way of teaching foreign languages based on grammar and translation led to what were called communicative approaches, which based the teaching on promoting the use of a language by inserting it as much as possible within natural communication situations (e.g., Howatt & Smith, 2014). The purpose of the teaching of a second and foreign language is that the student absorbs (via meaningful communicative experiences) the structures of that lan- guage, that is, she learns its grammar without even being aware of it. One could say that she learns her grammar implicitly (in the same way that one learns how to speak on the ontoge- netic plane). The concept of implicit grammar has therefore had an impact on teaching and it TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 7 In the teaching of students’ first languages, it could be assumed that boys and girls of school age dominate the basic structures of their language (i.e., they implicitly know the grammar of their language). Thus, the goal of grammatical teaching at the primary and secondary levels would be to learn basic grammatical concepts, that is, to learn explicit grammar that will serve as an instrument to improve formal uses such as in writing (see Rodríguez-Gonzalo, 2012). Nonetheless, although children in pre-primary and primary schools have an im- portant command of the basic structures of their language, studies of the develop- ment of language (see Serra et al., 2000) assert that there are complex structures (especially those of written and formal oral uses of the language) of late acquisition, and that these might not become dominant without the influence of schooling. No- tably, the seminal work of Lentin (1974) for nursery school and Camps (1986) for the first levels of primary school maintain that focusing on the use of complex syntactic structures necessary for the monologic use of the language such as written language is paramount. The need to teach students complex structures of language has not been limited to the first levels of schooling. Learning to write formal discursive genres requires the teaching of specific textual and syntactic structures as an essential component for their command (Zayas, 2012). In Spain, since the 1970s, procedures for implicit learning have become generalized as the exercise of a complex syntax typical of writ- ten language and mirroring the so-called “sentence combining” and the “phrase de base” in the Anglophone and Francophone areas (see Chartrand, 1996, and Connors, 2000). Just as in those areas some authors maintain that explicit teaching still is nec- essary (see Hudson 2007 and 2016, and Tisset, 2010), this is also the case in Spain (see Fontich & García-Folgado, 2018). Thus, since the teaching of implicit grammar also has its place in the teaching of the first language, the discussion does not arise has been possible to talk about the implicit teaching of grammar. Yet, as Ellis (2006) maintains, there are ways of teaching grammar that involve only practice or only repetition of structures of the target language: “First, some grammar lessons might consist of presentation by itself (i.e., without any practice), while others might entail only practice (i.e., no presentation). Sec- ond, grammar teaching can involve learners in discovering grammar rules for themselves (i.e., no presentation and no practice). Third, grammar teaching can be conducted simply by expos- ing learners to input contrived to provide multiple exemplars of the target structure.” (p. 84). A number of problems have been identified in the only-communicative learning of languages (e.g., Grenfell, 2000), which has led some researchers to re-consider the need to consciously learn the structures of the target language (e.g., Trévise, 2009). In 1990 R. Ellis was already advocating a double input in the teaching of languages: one focused on the form (that implies explicit knowledge and conscious concepts) parallel to the input focused on meaning (implying implicit knowledge, not aware of the usual expressions of the language) and maintaining that in the framework of a communicative approach promoting the conscious learning of the gram- mar was also necessary (Ellis, 1990). As DeKeiser (1998) states, “despite the numerous inves- tigations in the field of cognitive psychology, we do not have evidence that abstract concepts can be learned without being aware of them” (p. 34). This would also be the position of cogni- tive linguistics and socio-culturally oriented studies (see Lantolf & Poehner, 2014). 8 A. CAMPS & X. FONTICH as a dichotomy that excludes one or another grammar but in the relationship be- tween the two. A frequent presupposition with long-lasting effects in teaching grammar, already formulated in the 1970s in the French context by Peytard & Genouvrier (1970), is that teaching grammar consists of “making explicit the speakers’ internal system of rules which underpins all her concrete speech acts, i.e., to make explicit her implicit grammar” (p. 35). This assertion presupposes that between implicit grammar and explicit grammar there is a direct relationship and that, should the former become conscious, the explicit grammatical knowledge would crystallize. But this does not take into account what the history of linguistics shows (Harris & Taylor, 1989): Ex- plicit knowledge about languages and the different models that explain them are sociocultural constructions elaborated by those devoted to the study of languages. Explicit grammatical knowledge does not arise directly from awareness of the im- plicit grammar that the speaker dominates and that is manifested in use, but is me- diated by the scientific or pedagogical models developed. Thus, to approach knowledge of the relationship between implicit grammar and explicit grammatical knowledge, we suggest referring to a new conceptual frame- work: that of metalinguistic activity, which takes into account the human capacity for conceptualization and explicitness of observations and concepts about language and the languages. 2.3 Metalinguistic activity as a source of grammar learning We call metalinguistic activity the verbal or non-verbal activity that has, as referent, the oral or written language itself, which is taken as the focus of observation, reflec- tion and analysis. Humans, in parallel with the acquisition and development of lan- guage, acquire and develop the ability to speak about language and to develop knowledge about it (Taylor, 2000). Just as the language is acquired by participating in exchanges with the individuals who speak it, the capacity for observation, manip- ulation and reflection upon it is also developed in this interaction. The school plays an important role in this development (Gombert, 1990) since the learning of reading and writing implies that the language is externalized, becoming something observa- ble that can be manipulated and about which children constantly reflect during the course of school activities (see Roth et al., 1996; see the seminal work by Carter, 1990). The learning of grammar should have its roots in this possibility. We can say, then, that the ability to take language as an object of observation and comment is inherent in the same use of language. It is a human capacity that is based on the possibility of taking any object or process from the world as a focus of observation and analysis and to elaborate knowledge about it. The difficulty of doing so with the language is that it is both an instrument and a focus of attention (Camps, 1998; Taylor, 2000), the former being the most natural way to use language (as a means for accessing the “landscape”, i.e., meaning and communication; see Garton & Pratt, 1989). TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 9 a) The levels of metalinguistic activity Research and theoretical proposals on metalinguistic activity recognize that it emerges in different ways (Camps and Milian, 2000; Culioli, 1990; Gombert, 1990; Karmiloff-Smith, 1992). The work of Camps et al. (2000) on the interaction of stu- dents in collaborative writing work identifies three levels of explicit manifestation of the metalinguistic activity: (#1) metalinguistic activity that is made explicit through procedural and non-verbalized actions―i.e., through changes in the writing pro- posals without any explicit metalinguistic statement; (#2) explicit metalinguistic ac- tivity through metalinguistic statements, formulated in a common language, and (#3) explicit metalinguistic activity through statements formulated in specific grammati- cal language (see also Fontich, 2016; also Ribas et al. 2014). This distinction is crucial to understanding the possibilities and limitations of grammar teaching related to language uses. The learning of explicit school grammar, which would belong to the third level of metalinguistic activity, is not installed in a vacuum: Indeed, children are capable of a lot of reflection on language, conveyed with their own words and those of others, and may elaborate spontaneous concepts about language long before using specific grammatical metalanguage (Fontich, 2014; Milian, 2005). The research also shows that the grammatical concepts of the third level and the terms that are associated with them do not arise directly from the spontaneous statements. There is no smooth developmental transition between the different levels; rather there is a qualitative leap between levels (#1)-(#2) and level (#3) related to the systematization of knowledge that requires the formal teaching of grammatical concepts. To better understand this problem, Vygotsky’s (1987) dis- tinction between spontaneous and scientific concepts is useful. b) Spontaneous concepts and scientific concepts Spontaneous knowledge is formed in practical and concrete experience and is based on the immediately observable properties of the object. The spontaneous concepts are experience saturated, are closely linked to everyday contexts, and lack system- atic organization, as is reflected in an unsystematic, incomplete, and often erroneous verbalization. Language development is underpinned by interaction and involves the development of the ability to speak about the language that is learned and spoken (Taylor, 2000). In this process, speakers elaborate concepts about it in the same way that they elaborate concepts about any aspect of the world in which they live (see Miller, 2011). They are spontaneous concepts that develop without specific inten- tional teaching. Contact with the written language favours the expansion of more or less spontaneous concepts. Many of these concepts that, as noted, are developed through interaction, are loaded with knowledge that comes from learning at school what tradition has already made common: word, noun, phrase, text, etc. However, they are not systematic concepts and are closely related to the contexts of language use (see for instance Casas, 2014). Scientific knowledge represents the generalization of human experience as it has been set in science. Scientific concepts involve (a) systematization and (b) a higher level of abstraction, since they require the discovering of relationships between 10 A. CAMPS & X. FONTICH objects that are not directly perceptible (Miller, 2011). Both characteristics make these concepts operative when explaining reality, in a process that starts from reality itself to observe it, systematize the data collected, and establish some generaliza- tions. Such a process leads us in a round trip to return to reality, with a richer and more rigorous vision. These concepts are not learned spontaneously but require an intentional learning and, therefore, must be taught. They also require a model of reference, taking into account that, as stated above, a model is not reality or its por- trait, but a sociocultural construction that aims to explain it. Research on students’ grammatical concepts and writing highlights students’ ability to observe and talk about the language and make relevant observations (see Fisher, 2004; Gil & Bigas 2014; Myhill, 2000; Myhill et al., 2012; Rättyä, 2013; Watson & Newman, 2017). But they also show a frequent lack of systematic concepts and the close dependence of grammatical knowledge with regard to the contexts in which this knowledge has been learned. What we have said so far can serve as a starting point for understanding the process of abstraction that involves the learning of concepts (i.e., of grammatical concepts) since it requires the ability to discover features of the linguistic object that are often not directly perceptible. In synthesis, when we speak of metalinguistic activity, we refer to a linguistic (or procedural) activity that has as its referent the same language (its functions, its forms, and the relations between them). Applying Garton and Pratt’s (1989) meta- phor, metalinguistic activity involves making the window glass opaque so that we are aware of it instead of accessing the landscape beyond it. This activity can be carried out at different levels: it can simply be manipulative or it can also be verbalized, ei- ther in everyday words or via specific metalanguage. Thus, taking the Vygotskian ap- proach in this paper when referring to the process of conceptualization, we use the term “concepts” to refer either to spontaneous or scientific concepts. For example, a teacher’s grammar knowledge may emerge as a set of experience-saturated and unsystematic concepts (i.e., spontaneous concepts) but at other times might emerge as a set of well-articulated concepts (i.e., close to what Vygotsky names “scientific concepts”, and whose prototypical image would be a scientific theory). In the follow- ing section we briefly address the process of abstraction in learning grammar, which some authors refer to using the term “conceptualization” (i.e., the process of form- ing a concept or the result of such a process). This means that neither simple lan- guage use, nor verbalization, nor the use of concepts imply by themselves the knowledge of the system of relations that is the grammar of a language. Rather, both are necessary conditions for being able to build it, although a process of abstraction, scaffolded by the experts and targeting at building scientific concepts, is necessary (see Miller, 2011). We must now ask ourselves what repercussions this approach has for the teaching of languages. TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 11 2.4 The processes of abstraction in the teaching of grammar The learning of grammatical concepts necessary for the explicit knowledge of gram- mar as well as for this knowledge to become operational in the use of the language, must start from observation of the behaviour of language forms in sentences or dis- course (e.g., corpora prepared previously by the students themselves or by the teacher, students’ texts, sets of sentences explicitly prepared to focus on a specific grammatical question, real headlines, etc.). Observation involves the perception of characteristics and comparison, which enables the detecting of similarities and dif- ferences, for example between verbs and nouns that indicate or refer to movement (i.e., event nouns such as Spanish “salir/salida” to exit/exit, etc.). This first process takes us to a first level of abstraction that has been reached by identifying the characteristics of the object and the contrast with other elements supposedly belonging to the same category (see Barth, 2004). Successive verifica- tions will lead to the generalization of the conclusions; for example, students may see that belonging to a lexical category (e.g., a noun, a verb) is not a matter of all or nothing, but rather that there is a gradation. Thus, while some elements have all the characteristics of that category (as prototypical elements) others will only share some of them (e.g., event nouns as non-prototypical nouns). The last step will be that of systematization, in other words integrating the category built into a system of relationships that gives it meaning (e.g., notions of verb-arguments will explain the behaviour of event-nouns with regard to regular nouns). This is regarded as a stepping stone for critical thinking within the framework of grammar (see for in- stance Janks, 2010; Schicker, 2018). 2.5 The teaching of grammar from the consideration of metalinguistic activity From the viewpoint of metalinguistic activity, the teaching of grammar should take into account: 1) the need to teach students complex syntactic structures of a de- ployed syntax especially necessary for command of the formal (particularly written) uses of the language; 2) the need to install metalinguistic activity as a centre of lan- guage teaching, and 3) the need to scaffold the students’ processes of abstraction as required by the learning of explicit grammar. These three aspects can be approached independently of each other, but they weave a web of relationships that may give coherence to grammatical teaching (see Fontich & Camps, 2014).5 In the first place, although the repetition by simple imitation without reflection can occur in the learning of languages, the teaching of complex syntactic structures can hardly be carried out without being aware of them. For example, learning to 5 Also, with respect to L2, Ellis (2010) considers that in form-focused instruction the character- istic that emerges as especially noteworthy is “metalinguistic activity involving such instruc- tional strategies as providing learners with metalinguistic information [...], inviting them to discover grammatical rules for themselves, and encouraging reflection on and self-repair of their errors” (p. 452). 12 A. CAMPS & X. FONTICH combine sentences to form a complex statement requires comparing, deciding what is eliminated, what is replaced by another form, and so on. The verbal activity on the forms of the language―the metalinguistic activity―almost always becomes indis- pensable. At school ages, this reflection is often accompanied by a specific metalan- guage that is being constructed, at the same time, through systematic teaching. We see then that the metalinguistic activity accompanies the learning of the uses of the language. A second consideration is that the metalinguistic activity is also the axis of the learning of systematized grammatical concepts. The abstraction starts from the per- ception and observation of elements, and comparison among them, which allows for the formulating of hypotheses that have to be verified. This process involves met- alinguistic activity in the form of peer- and student-teacher interaction (see the stud- ies in Ribas et al., 2014); that is, learning concepts only through rote-learning or by formulaic exercises might not be possible. Thus, the metalinguistic activity (as the axis and purpose of grammatical teach- ing) is installed as the centre of the learning activity of grammar with regard to both implicit and explicit grammar. Also, as pointed out a few years ago (Camps, 2014; Milian, 2014), it is necessary to focus grammar teaching on two spaces, that of lan- guage use and that of systematization6. The challenge is to weave such intermediate space by means of a teaching style that takes into account students’ ability to carry out a metalinguistic activity that, with the teacher’s scaffold, enables a process of shared construction of knowledge. 3. METHODOLOGY Since the relationship between grammar and writing is questioned and the focus is placed on teachers’ thinking and practices, the goal of the study is to explore teach- ers’ concepts of grammar, grammar teaching, and grammar-for-writing. A working hypothesis is that the teachers’ concepts of the two fields (grammar and writing) are two-fold: they can be an obstacle that hinders the necessary renewal of grammar- for-writing practices, and at the same time can offer anchoring points for this re- newal. Our assumption is that the latter can be substantiated through the reflective activities learners are engaged with (and which we group around the category of “metalinguistic activity”, see above) rather than on the basis of the implicit/explicit knowledge dichotomy. The project comprised three main actions: a Likert questionnaire, a set of semi- structured interviews, and a cycle of seminars sustained overtime. They were accom- panied by punctual actions: the organization of two local conferences where partic- ipants were invited to take part either as audience or as speakers, and the 6 This resonates with Ellis (2006) who asserts that “grammar instruction should take the form of separate grammar lessons (a focus-on-forms approach) and should also be integrated into communicative activities (a focus-on-form approach)” (p. 103). TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 13 contribution to a special issue. The Likert served to select a set of six teachers, as representative as possible of different sensitivities towards grammar teaching, who were to be engaged in a cycle of seminars and to be interviewed before and after it. The objective of the first interview was to explore the problems that, in the teacher’s opinion, were present in the writing-grammar interplay, in order to put them on a common footing, and to address possible solutions at the subsequent cy- cle of seminars in which the six interviewed teachers and four researchers would participate; the second interview aimed at a final reflection about the whole project. This was based on the idea that teachers’ concepts could undergo a process of change when dealt with overtime and within a community of discussion, and that this was a fundamental step when exploring paths aimed at teaching innovation (Bastons et al., 2017; Engeström, 2011; Fontich & Birello, 2015). Seminars are cur- rently under exploration (see Fontich, 2017) and will not be addressed in here. In the present paper we present some results of the Likert questionnaire and then exten- sively focus on the first interview with one of the teachers. Likert questionnaire With regard to the Likert questionnaire, a cohort of Primary and Secondary teachers (n = 94) was established (details of the sampling process are shown in the Appendix) and a cluster analysis was adopted for grouping the participants according to simi- larities and dissimilarities in their responses. The objective was to explore their de- clared teaching practice in teaching grammar according to the following six variables: (1) Memorization, (2) Metalinguistic writing, (3) Oral linguistic variation, (4) Student participation, (5) Transmission, and (6) Interaction. Answers were measured on an 11-point scale (0 = hardly ever and 10 = very often) and they were meant as far as possible to select a representative cohort (n = 6) from the differing tendencies among the respondents. Interviews Each teacher of the cohort was interviewed twice, in 50-minute sessions that were held on school premises. Semi-structured interviews focused around three issues at stake: grammar; grammar-writing instruction; and grammar-writing learning. In a brief and informal introductory researcher-practitioner meeting, the interview pro- cedure and the adjective “semi-structured” were explained, and the practitioner was also provided with the aforementioned list of items (Guasch & Ribas, 2013). This list of contents and the list of basic procedures for the interview (e.g., respecting the interviewee’s moves to shift the focus of attention or alternatively changing it when- ever the conversation became strained) were elaborated by the group of researchers after a thorough discussion. The transcription was carried out based on broad criteria that reflected the basic features of the interaction and with the interest placed on the content, without going 14 A. CAMPS & X. FONTICH into detail of prosodic, gestural aspects, etc. (Llisterri, 1996 and Payrató, 1995; we numerate each turn, use “mmm:” when they take their time to think and indicate with “(...)” whenever some parts have been withdrawn for exposure clarity). It was performed by a person outside the investigation and subsequently reviewed by two researchers, one of whom was the interviewer. The analysis of the interview was carried out via an iterative process based on periodic meetings of the research group (see Cubero et al., 2008; Guasch & Ribas, 2013). The procedure for the analysis followed two steps within a general content analysis approach (Guasch & Ribas, 2013). The first step was an inductive oriented sequential analysis, cutting into segments the different raw themes identified by re- searchers as they appeared chronologically. To compare the themes identified among transcripts each segment was given a title, an indication of whose were the opening turns, and a short description. This measure facilitated the discussion with the rest of the members of the research group, and was aimed at reaching a high level of agreement with regard to the second step of the analysis, a deductive-ori- ented conceptual analysis focusing on grammar, grammar-writing instruction, and grammar-writing learning. From these concepts, which appeared more or less explic- itly throughout the interview, we observed the indirect emergence of various ideas that could be grouped around the category of “metalinguistic activity”, and that we considered were especially relevant to supporting the subsequent work of the sem- inar. The result of the syntagmatic analysis is presented only briefly, since the pre- sent article focuses on the concepts of the teacher regarding the aforementioned topics, i.e., on the categories in which the paradigmatic analysis is articulated. 4. RESULTS OF THE LIKERT QUESTIONNAIRE Table 1 presents the variables, the statements, the final average score, and the score dispersion/concentration. The results showed teachers’ slight tendency to adhere to what research deems as good practices in variables (1) to (4), and a slight tendency to adhere to what research deems as bad practices in variables (5) and (6), being the answers in all variables (but (4) and (5)) dispersed (see Bastons et al., 2017, and Fon- tich & Birello, 2015 and for some deliberations on the rest of the Likert). TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 15 Table 1: Teachers’ declared practice Variable Statement Score* Re- Score*: disper- Adhesion to deemed sults: sion/concentration good/bad as good Aver- practices practice age Score * (1) Memo- “I devote time to Low 4.2 Dispersion: rization memorizing gram- score • 0 to 3: 45% mar rules (on syn- • 4 to 6: 35% tax, morphology, • 7 to 10: 20% orthography, etc.).” (2) “I finish my expla- High 5.2 Dispersion: Metalin- nations by asking score • 0 to 3: 30% guistic writ- the students to • 4 to 6: 25% ing sum up through a • 7 to 10: 45% metalinguistic writing the gram- mar phenomenon we have worked upon.” (3) “I refer to oral High 5.4 Dispersion: Teachers’ Oral lin- (colloquial expres- score • 0 to 3: 20% tendency to guistic vari- sions, registers on • 4 to 6: 40% adhere to ation TV and radio, etc.) • 7 to 10: 40% what re- as a resource to search deems reflect upon as good prac- grammar phe- tices nomena.” (4) “I base my work High 7.2 Concentration: Student on problems score • 0 to 3: 10% participa- found in students’ • 4 to 6: 20% tion productions (e.g. • 7 to 10: 70% related to mor- phology, syntax, orthography, punctuation, etc.).” (5) “I present activi- Medium 7.8 High concentration: Transmis- ties that require score • 0 to 3: 10% sion application of • 4 to 6: 10% Teachers’ grammar content • 7 to 10: 80% tendency to as previously in- adhere to troduced.” what re- (6) “I set up students High 3.8 Dispersion: search deems Interaction in small groups to score • 0 to 3: 45% as bad prac- reflect on lan- • 4 to 6: 35% tices guage phenom- • 7 to 10: 20% ena.” *Point scale: 0=hardly ever and 10=very often. 16 A. CAMPS & X. FONTICH Teachers discard memory-related procedures (in (1)). This is coherent with defend- ing (in (2)) a reflexive use of the writing (55%), i.e., a metalinguistic writing, in order to synthesize the explanations made during the lesson, as well as the use (in (3)) of oral expression (TV, radio, oral colloquial...) to reflect on the grammar (over 50%). In accordance with this interpretation, teachers could be inclined to work in a more inductive way, starting from the problems detected in the texts. However, Fontich and Birello’s (2015) analysis shows that for some teachers the “inductive approach” means “without necessarily a previous plan of intervention”: only a low number of teachers (less than 9% of the sample) choose planning guidelines in text revision as a good practice, a result consistent with the difficulties identified by the research in terms of evaluation in the written composition process (MacArthur, 2013). Also, stu- dent participation (4) stands out with a more than 80% positive rating (points from 7 to 10). This answer would agree with a general tendency described in Fontich & Birello (2015), according to which, in order to learn grammar, we must reflect on the linguistic forms and their functioning in the discourse and know how to explain this function. Contrariwise, we can also observe a positive high agreement around 90% with regards to transmission (5), considered by the designers of the Likert as a bad prac- tice. It can be interpreted that this statement responds to a basic intuition among teachers, according to which procedural knowledge derives necessarily from a pre- viously-worked declarative knowledge. This is a deductive schema whose productiv- ity is strongly questioned by experiences based on the students’ inquiry (see Rodríguez-Gonzalo, 2015; Simard et al., 2010), but is strongly defended in Spain by influential linguists (e.g., Bosque & Gallego, 2016). In this sense, this answer would be consistent with the answers in #6 on the benefits of grammatical discussion in small groups, which we consider good practice but which more than 50% locate in the negative range. This disparity has been identified in empirical works that show the difficulty of clearly identifying the items to be taught and the need for the teacher to provide a clear framework in which to place them (a phenomenon that is referred to in some of the works in Ribas et al., 2014 and Dolz & Simard, 2009). Through a cluster analysis, participants were classified according to similarities and dissimilarities in their responses. Table 2 identifies the variables that distin- guished the different groups and Table 3 shows that groups 1 and 4 were considered closer than groups 2 and 3-to those practices deemed by the researchers as good or bad. TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 17 Table 2: Statistical description of declared practice per each group Groups Means (1) (2) (3) (4) (5) (6) and Sd Memori- Meta- Oral ling. Student Trans- Interac- zation ling. writ- variation particip. mission tion ing Group 1 mean -,72 -,52 ,07 ,30 -,85 ,73 Sd ,48 ,81 1,04 ,81 1,17 ,75 Group 2 mean -,17 -,98 -,23 -,69 -,16 -,90 Sd ,74 ,60 1,02 ,97 1,00 ,37 Group 3 mean ,34 ,82 -,43 -,41 ,55 -,70 Sd 1,30 ,55 ,93 1,02 ,62 ,52 Group 4 mean ,48 ,73 ,47 ,69 ,41 ,77 Sd ,95 ,52 ,82 ,51 ,50 ,74 Total mean ,00 ,00 ,00 ,00 ,00 ,00 sample Sd 1,00 1,00 1,00 1,00 1,00 1,00 Table 3: Profile of each group according to practice deemed as good Strong varia- Transmis- participa- Weak variables variation Memori- Metalin- Oral lin- Student Interac- bles of good ++ Groups writing guistic guistic zation of good + and and bad -- sion tion tion bad - practice practice 1 ++ - ++ 2++ 1- 2 - -- -- 2 -- 1- 3 + -- -- 2 -- 1+ 4 + + ++ -- ++ 2++ 1 -- 2+ Six teachers were contacted out of the four groups identified in the Likert question- naire and were invited to take part in a cycle of actions within a formative period of eighteen months. They all accepted and Table 4 below presents the profile of each of the six. 18 A. CAMPS & X. FONTICH Table 4: Profiles of the six teachers that accepted to take part in the cycle of seminars and in- terviews Gender Degree Language Modality7 Group Years School immi- of instruc. teaching grat. rate Ma- Female Primary Catalan Primary 1 <10 Medium ria* educat. Spanish English Cesca Female Primary Catalan Primary 2 >25 Medium educat. Spanish Vera Female Primary Catalan Secondary 3 10 to 20 Medium educat. Lang. Arts Sara Female Lang. Arts Catalan Secondary 1 <10 Low Carla Female Lang. Arts Catalan Secondary 4 >30 High * Names are invented. 5. RESULTS OF THE ANALYSIS OF AN INTERVIEW WITH ONE OF THE TEACHERS: CARLA We include here the initial interview with Carla, a secondary school teacher selected on the basis of her long professional career (she was the oldest teacher) and her professional profile (according to the Likert profiles she was placed in group 4: strong accent on interaction and student participation). In the interview, as will be seen, she has a strong position on an approach to grammar anchored on the idea of the explicit/implicit grammar knowledge dichotomy, while at the same time she tenta- tively considers alternative ways of dealing with grammar by engaging students in reflective practices sustained over time. The interview was carried out in 569 turns. Carla revealed that in all her long professional career she had never been given the opportunity to express her thoughts and feelings, let alone to do so in the company of other primary school teachers, and profoundly appreciated participating in these seminars. During the conversation she was very active. On the one hand, the number of words of Carla’s interventions (5,397) exceeded those of the interviewer (4,299) and, on the other hand, Carla selected and very often introduced the topics she was interested in 7 Although we were interested in keeping a balance between Primary and Secondary partici- pants, we did not establish how many teachers in groups 1 to 4 belonged to each modality. Instead, possible differences between Primary teachers and the general sample were explored through a three-fold multiple choice questionnaire about (a) intervention strategies, (b) as- signments for text production, and (c) error correction. Interestingly, results suggested that Primary teachers were closer to what is considered good practice, although strong tendencies could not be identified. TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 19 speaking about (15 times compared to 12 by the interviewer). Agreement between the research team through a process of triangulation of this first segmentation ena- bled the dividing of the interview into 23 segments. Initially Carla placed her focus on grammar (segments 1-6), establishing her own diagnosis (its uselessness for improve writing), school courses (not taught well), and the failure of concrete proposals (problems are repeated every year). Diverse topics were addressed as the conversation progressed and the interviewer explicitly main- tained the focus on the grammar-writing relationship (segments 7, 13, 17 and 21). Some of Carla’s focal points included: her vision of what constituted grammar (seg- ment 8); the socio-cultural changes she observed in students (segment 9); the im- portance of grammar in learning normative aspects (segment 10); how the use of electronic devices constituted an obstacle to this relationship (segments 11 and 12); the need for guidelines to direct students and at the same time their inefficiency in doing so (segments 14 to 16); and/or the lack of grammatical instruction in elemen- tary school as a possible cause of the problems found at secondary levels (segments 18 to 20). Carla considered that her positions were widely shared by other language teachers and that this endorsed the reliability of her position. She accepted that two possible ways of overcoming the situation might be to integrate the grammatical teaching of the languages taught and to work on writing in all areas (segments 22 and 23). However, this highlighted some additional obstacles: teacher training; the inability of some teachers to connect with the new generations; the refusal of faculty members to work with the language in all areas; and the obsolescence of the sylla- bus. 5.1 The conceptualization of grammar In Carla's discourse two concepts of grammar emerged, which seemed simply to co- exist as independent realms because she did not relate one to the other: grammar as implicit knowledge (implicit grammar), and grammar as explicit, declarative knowledge (explicit grammar). Although she never named them in this way, she did describe them and reflected on the students’ difficulties in relating declarative knowledge with writing. 1) Implicit grammar On the one hand Carla considered that knowing grammar was knowing how to use language, without necessarily being aware of this. In this sense, the implicit grammatical knowledge manifested itself in verbal production and comprehen- sion. This is clear from her words, which suggested that this first concept of grammar coincided with that of linguistic competence (168. there are people who are illiterate in the sense that they have never been to school, they do not know how to read or write and yet they do know how to construct sentences correctly). This grammar is innate and does not even imply knowing what gram- mar means; it is, therefore, unconscious (206. then it is that to see the grammar is innate, that is you must have it to be able to communicate even if you do not 20 A. CAMPS & X. FONTICH even know what grammar means). Also, grammar is related to thought and its structure. Carla’s insistence on this interrelation was strong, although it was fraught with ambiguities and some contradictions. According to Carla “thought is language” (162) and grammar is the structure of thought (168. the structure of what you think of your thinking is grammar). An interesting statement (still associated with the idea of grammar as a struc- turing of thought) was the evocation of those slower “old days” (i.e., herself as a child or, generally speaking, before computers and the Internet), which al- lowed language reflection while reading or writing. Two autobiographical anec- dotes explain this idea well (206, 208, 220, 224): as a child Carla used to notice words, sentences, labels, titles, etc. that had some grammatical incongruity (see below “5.4. Metalinguistic activity”). This set of characteristics led her to affirm that this grammar was not learned in class (24. then the student who knows [...] how to write, who knows how to think, who knows how to structure her thinking, mmm: I think that it is not the result of what she has learned in grammar lessons). She insisted that thought was grammar, suggesting that grammar knowledge was inherent in the very experi- ence of learning the language (24. you have it if you are in contact with people). The idea she clearly explained was that linguistic competence, the mastery of grammar, could be improved via reading and listening to stories. In this case, it would be an implicit learning: 296. if you have a student who does not open a book it is difficult not to open a book, that does not listen to stories, that does not work, that does not exercise this muscula- ture of the reflection, it is very difficult in my opinion for her to write a competent text, that she can dominate the grammar About this relationship between grammar and thought Carla’s starting point was that the poverty of students’ linguistic repertoire was one of the barriers for reflection in general and specifically for reflection on the language (although she watered this down by saying: 136. maybe at the end of the day it is all the same because thought is thought, but I reckon that words, or reasoning through the word, is not so common anymore). 2) Explicit grammar Carla's discourse about the grammar that was taught or should be taught in school was, in some ways, rather confusing. Two ideas articulated her discourse on this topic. First, the main function of teaching grammar was the knowledge of the rules of grammar, which were the basis for writing well (296. the baseline for good writing is in grammar norms [whether] you like it or not). The central axis of her discourse when referring to grammar and writing was the correction associated with these rules. All the examples she gave referred to orthography (262, 280) but we can infer that these norms could also refer to morphology or syntax (44). Secondly, the other idea that emerged, albeit less clearly, was the place that conceptual knowledge occupies or should occupy. It can be inferred that for TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 21 Carla to know a concept served for having a good command of the norm and for correction. She explained two types of problems. On the one hand, the students might not understand the concepts and integrate them in language use (44. you come to the conclusion that they have not understood what they are supposed to understand, you know what I mean, that is, if you teach them a rule to apply for instance when you find a direct object do not introduce it with a preposition and they keep inserting the preposition). On the other hand, some concepts, es- pecially those related to the textual grammar might not have been taught suffi- ciently (352. when the student is told that connectors fail, she must know very well what a connector is and many times you realize that work about connectors has been scarce or non-existent). 3) The relationship between the explicit knowledge of grammar and writing The two outstanding ideas come together in the statement that Carla repeatedly formulated: students did not integrate conceptual knowledge and, therefore, were not capable of using it as a tool to write correctly. She recognized that students sometimes appeared to have learnt the grammar they were taught and were able to pass an exam, but were unable to apply such knowledge when writing (9. we examine them and mmm: well, there is then that they do well and learn at that moment what they are taught but then when putting it into practice they keep making the same mistakes). Her words were forceful when the interviewer raised, among others, the topic of the relationship between writing and learning the grammar: 6. look at me, after so many years of mmm: devoting time to it and thinking it over I have come to the conclusion that I have no solution, no solution for the main problem, which is divorce between writing and learning grammar, a remarkable divorce Due to the profound writing-grammar divorce the pupils apparently learnt the content they were taught (6. every year the same thing, repetition of the gram- matical contents) to pass the exams but they were not capable of translating this conceptual knowledge into practical knowledge and kept making the same mistakes in writing over and over again. Thus, for Carla concepts and practice were placed in two different worlds. Students saw grammar content as just a subject (15), detached from reality and in the pejorative sense that can be given to this term. These overwhelming initial statements were argued throughout the interview and Carla discussed their possible causes. In spite of this, she did not abandon the idea that in order to write better one needed to have some declarative knowledge of a conceptual type that was essential in the revision of the text (352). Nonetheless, in her initial conception, the idea of grammar of the text or discourse did not appear (she referred to grammar in a very broad sense, e.g., in 288). She only talked about it when asked directly about the topic by the in- terviewer, relating it to the correctness of the texts (e.g., to avoid repetitions, in turn 80) and, towards the end, to the use of connectors (352). Grammar 22 A. CAMPS & X. FONTICH (referred to as a basic knowledge about language) would be a sine qua non for good writing (294. I think it is inherent, that is, you can hardly make a correct construction if you do not have the base). To finish we could say that Carla believed that students should be able to estab- lish a relationship between the grammar taught, the rules, and the practice of writing. She also believed that one of the causes of the inability to do so was a cultural one, especially the lack of time to think caused by the speed inherent in the new technological era. 5.2 The conceptualization of teaching of grammar As was natural, Carla’s concepts of the teaching and learning of grammar were closely related to her concept of the nature of grammar and mediated by her con- cepts of teaching and learning. In the previous section the relationship between the concepts of grammar and its teaching was identified. In this section this will be com- pleted by focussing directly on the concepts of teaching and learning. The first thing to note is that Carla spoke of teaching (12 times) and learning (17 times) while the interviewer put the focus on teaching (13 times, compared to 2 ref- erences to learning). This fact is relevant, especially because Carla established a con- flictive relationship between what was taught (and what was not taught) and what was learned (and what was not learned). The first idea that appeared about teaching and learning grammar was that of failure, especially in relation to the incidence of grammatical knowledge in the improvement of writing. How did Carla conceive the teaching and the learning of grammar? Two main ideas emerged from her words. First, the teaching of grammar was explicit and trans- missive; it was taught through teacher explanations and through the correction of written essays. Now, this type of teaching did not trigger learning as a result of con- ceptual understanding. Rather, it triggered rote learning that did not make sense to students. Not even those who learnt the “theory” (36) were capable of using this knowledge that did not even allow them to apply the norm (36. they can [make] a study that is a direct complement, and do it a thousand times later, do it well, learn it, the theory, learn that from memory perhaps, and then make the same mistake fifty thousand times (...) I do not know why). Secondly, the other idea was remarkable for its impact on the conception of the relationship between teaching and learning grammar and writing. It was related to Carla's concept of grammar, understood as the structure of thought. According to her, this grammar was implicit, inherent in the use of language; it was not taught, it was learned implicitly through sustained reading and thinking and contact with the written language and other speakers. In short, the grammar taught (explicit) was not learned or learned poorly, and that which was learned (implicit), which was the basis of the use of the language, was not taught. Moreover, prevailing social and cultural contexts prevented TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 23 learning: language and the capacity for reflection were impoverished by the increas- ing rapidity demanded and facilitated by the new technological methods. 5.3 Conceptualization of writing and of teaching-and-learning-to-write The concepts expressed by Carla about what it was to write, and about teaching and learning to write, were closely related to those she expressed about grammar and its teaching and learning. First of all, according to her, writing was knowledge that one possessed or did not possess and that was related to the ability to think, and to structure thought: 26. the student who can write it’s not because of grammar, but because she knows, because she knows how to structure her thinking, and, on the other hand, the students who do not mmm: may not have sufficiently exercised this, they have not done these gymnastics at the time of translating this onto a piece of paper and then it is difficult for them. Writing, according to this, implied searching calmly within oneself for the ideas that would be expressed through the words (308. therefore, they need tranquillity so that they can go and search within their inner files, finding this tranquillity to get the words is hard work for them). The ability to think and structure thought was associ- ated with slowness, which offered the opportunity to think and reflect in order to learn (152. it stays inside you). This was not possible with the use of the Internet, which, due to the speed with which it usually functioned, did not allow one’s own thinking to centre on the process of writing: 153. and for example, it happens to me that if I read a book for example, slowly, that is to say if I read fairly slowly, my thought keeps reflecting, and it has time, to reflect and understand, and whenever you get to understand something, it stays inside you, whereas if I surf on the Internet everything goes so fast, I cannot retain half of the infor- mation. The idea of calm also appeared when describing a situation related to writing activi- ties in the classroom. When the students were asked to write a text, she encountered reluctance among some of them, especially the more restless ones who expressed their difficulty in finding ideas to write about. To be able to do it, they needed calm and some training: 308. it takes like ten minutes to calm them down mmm: that is to say they have their heads all over the place, and therefore they need calm, then in order for them to be able to look for that tranquillity within their inner files, to find the word, it is much harder for them than for those students who already have it trained. Yet, despite the previous assertion referring to the need for training, Carla did not associate it with teaching in the classroom. Only at the request of the interviewer did she recognize that one can teach how to compose a piece of writing, but she maintained that it should be done before the middle level of secondary school (14- 16 years of age) (344, 346). In response to a direct question from the interviewer 24 A. CAMPS & X. FONTICH Carla also suggested two practices that could help to fuel reflection within the pro- cess of writing: reading of model texts (408), and comparison of texts (412). In fact, Carla’s predominant idea about writing was related to that of correction. The learning of writing pivots on orthographic and syntactic rules. When talking about the concepts of grammar and its teaching (see above), we highlighted Carla’s belief in the importance of explicit knowledge about the norm. She considered, then, that the school’s task was to teach students the norm for writing, mainly the ortho- graphic one. This was the purpose of the exercises and activities that were carried out (254. because we have to learn to write correctly). The usual practices were the explanations of the teacher and the correction of the written texts. While the term “falta” (grammatical or orthographical error) referring to spelling and syntactic rules appeared repeatedly in her interventions (e.g., turns 32, 40, 256, 280, 324, 328, and 436), she used the term “error” (mistake, less tied to grammar and norm and closer to communication) for referring to textual problems (414, 420). In short, for Carla to write was to express thought, embedding it in writing through words, according to the norm. While the expression of thought was learned by impregnation, the rules were learned in school through lessons and exercises and the correction of the writings. And while the objective of education was to teach these rules so that students could learn them and write correctly, the ability to ex- press thought could not be taught. It is noteworthy that the divorce between writing and grammar detected by Carla also appeared in the same idea of writing as an ex- pression of thought on the one hand and as a normative correction on the other. 5.4 Metalinguistic activity So far, analysis of Carla’s concepts has revealed a dichotomous vision between writ- ing and grammar, and between implicit grammar and explicit grammar. These di- chotomies were also reflected in her explicit approaches to teaching: the object of teaching writing and grammar was restricted to the contents, which had an impact on the domain of the norm, while reflections on the need to teach writing texts arose in all cases in response to the interviewer’s questions. However, some very interest- ing ideas related to what we call metalinguistic activity did appear spontaneously. In the first place, two anecdotes from her own childhood suggested to us that Carla recognized how children are, from a very early age, capable of reflecting on the language, of taking it as an object of observation: 206. but at that age I already needed to, know the grammar even though I did not know it was called grammar, and I remember that my mother bought me a story, a book of stories that was called “The one thousand and one nights” [i.e., The Arabian Nights] and I used to ask my mother why is it called “the one thousand and one nights”, and my mother did not understand the question (...) and I wanted to ask her why do they write “one nights”? I could notice that there was something in there, and I had no idea of what the grammar was or what the language was or anything. TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 25 Carla, as a child, adopted a linear stance on the noun phrase of the title and inter- preted the word “one” as if it were an article: it should then have agreed with the noun “nights”, but it did not. Indeed, as a child her interpretation was not correct, but Carla saw it as an example that showed how the little girl was able (spontane- ously) to carry out a metalinguistic activity on a certain grammatical issue. The same thing happened in the other example that Carla provided, which showed the capacity to judge the adequacy of the nouns derived from verbs and the lack of regular appli- cation of a rule (224. there was a garage near where I lived with a shop-sign stating “lavado y engrase” [washing and oiling] and I always used to think why do they write “engrase”? because I should think they were meant to say “engrasado” instead). 8 Although it is not obvious, we can consider that one of the meanings given by Carla to grammar (a tool to structure thought) would make it possible to reflect on written texts. This seems to be inferred from some of her previous words: 24. then, the student who knows how to write, who knows how to think, who knows how to structure her thinking, mmm: I think that it is not the result of what she has learned in grammar lessons. That is, reflection was the basis for learning to write and also for learning grammar. However, considerations about children’s reflective abilities (sometimes trained through reading) did not lead to Carla spontaneously raising the need for activities to promote reflection as a basis for grammar school learning. As noted, she at- tributed these skills to some students, as something idiosyncratic that might have been developed outside the school or before middle secondary education where she teaches Language Arts. Nonetheless, when, at the interviewer’s request, she imagined some activities for learning grammar in relation to writing, she described possible tasks that would involve grammatical reflection, such as the imitation of text models or the compari- son of texts (408. look, I do not know how to answer this, but we can first, for exam- ple, read and explore a text that is correct, and then give the possibility of three texts so that they have to demonstrate why one is better than another). Carla considered that activities of this kind allowed us to reflect on the errors of the texts or on the linguistic resources necessary to express the intended meaning. She gave an exam- ple of a frequent problem in the students’ narratives (414. is that you tell them mmm: explain what you did this morning, and they go like I got up and had breakfast, I have left home, I have taken the bus, etcetera, you see that they change the verbal tense with notorious ease), and concluded that to solve it required the teacher’s help (i.e., it needed to be adequately taught): 8 In the original Spanish the expression is “lavado y engrase” and Carla wondered about “lavado” (as a regular participle from the infinitive “lavar”, wash) but “engrase” (which acts as a noun that does not stem from the verb “engrasar”, whose participle is “engrasado”). 26 A. CAMPS & X. FONTICH 418. well you should teach this to the student, and you have to tell her “look, verb tenses express past present and future, and when we talk about something in the past, we can see that the past can be more in the past or less so in the past. Thus, beyond the dichotomy between grammatical learning and text writing that we could consider as her strong thinking, Carla caused ideas to emerge that could un- derpin a reflective learning of grammar based on a conscious teaching of linguistic forms in relation to texts. These ideas could function both as a change regarding tra- ditional conceptions about grammar and writing and as a focus on metalinguistic activity. The latter would become the actual objective of grammatical teaching, serv- ing as an instrument of reflection both on the grammar system and on the uses of the language. 6. CONCLUSIONS The results of our analysis suggest the complexity of teachers’ concepts, consistent with what has been highlighted by studies on teaching grammar and writing (see Watson 2015a and 2015b). A first observation that we can mention relates to the profile in which the teacher interviewed, Carla, was initially grouped according to the Likert questionnaire (group 4: strong accent in the interaction and participation of the students). The results show that this profile is blurred in the context of sustained discussion over time. A Likert questionnaire completed in isolation may well lead to a sort of fixed photograph that, regardless of how accurate it might be, cannot fully describe the complexity of a practitioner’s thinking. This has been confirmed in the analysis of the interview to another teacher par- ticipating in the project as well as in the analysis of a part of the seminars (Fontich & Camps 2015; Fontich, 2017), and is consistent with the sociocultural approaches that consider the construction of knowledge as a result of sustained participation in dis- course communities (see Mercer, 2013). This participation would trigger a process of transformation and re-construction of knowledge. In this sense, diverse studies have described the transformation of students' knowledge into situations of interaction (both in the classroom situation and in in- terviews with the researcher outside the classroom) and in various areas of the cur- riculum (e.g., Schwarz et al., 2009) including grammar and grammar-writing (e.g. Casas, 2014; Myhill et al., 2016). What has been highlighted as a key aspect of these studies is that the research design itself catalyses the reflective process of the par- ticipating subjects through the interaction with the researcher or with the other par- ticipants (see Camps, 2000). Engeström (2011) argues that this phenomenon actually constitutes a programmatic pillar of social science research through what the author calls “formative interventions”. In this sense, we could say that the interview con- ducted stimulates the reflection of Carla, our teacher, and makes a complex thinking emerge, located in a set of dichotomies. TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 27 Carla’s thinking about grammar and its teaching and its relation to writing and its teaching was expressed more or less explicitly as dichotomies that were related to one another (Table 5): Table 5: Dichotomous concepts in Carla’s discourse Implicit grammar: unconscious (“linguistic Explicit grammar: concepts, rules, and norms competence”) Writing as an expression of “what you have Writing as a process of adjustment to a nor- inside” mative Learning by impregnation via the activity of Learning via lessons and correction of writ- reading ings These dichotomous conceptions that infuse habitual practices can constitute real obstacles to advancement in a grammar teaching that may have an impact on the use of the language. Despite the fact that in her discourse such dichotomies (that we express as “divorce”) appeared repeatedly, indirectly there was also the idea of re- flection and that young children were capable of making very insightful reflections on the language. This is what we have categorized as “metalinguistic activity” and the fact that Carla accepts this could be the basis of effective teaching. In sum, Carla’s conceptualizations can be clearly seen as installed on the explicit-implicit dichotomy, yet she also refers to the need of engaging students in reflexive practices. We inter- pret this as a vague awareness of the need to promote metalinguistic activity. This reference can be interpreted as a potential anchor to shift the focus towards foster- ing reflection, as a source for grammar learning and instead of just direct instruction. A number of studies (e.g., Fontich, 2016; Myhill et al., 2012; Rättyä, 2013; Ribas et al., 2014, etc.) consider that for this shift to happen, assistance might be provided to teachers so that they can (i) crystallize their intuitions about the relevance of met- alinguistic activity within the grammar-writing interplay, and (ii) design cycles of ac- tivities devoted to engaging students in this activity. However, this study also sug- gests that there are major difficulties when trying to translate outcomes of research conducted “from within school” to a population of teachers who are not directly fa- miliar with research. Furthermore, while engaging teachers in valuable processes of reflection is seen as the crux of the matter for innovation (Engeström, 2011), the results of the analysis presented suggest that such a reflection needs to be sustained over a long period of time. As we have noted above, a working hypothesis is that the teachers' concepts can be an obstacle that hinders the necessary renewal of grammar-for-writing practices, and at the same time can offer anchoring points for this renewal. We consider that the latter can be substantiated through the reflective activities teachers engage the students in (and which we group around the category of “metalinguistic activity”, see above) rather than on the basis of the dichotomy of the implicit / explicit knowledge. The objective of involving students in reflective practices is based on the idea that activity is the framework in which learning occurs. This activity must be 28 A. CAMPS & X. FONTICH oriented by the teacher towards a purpose properly explicit and should arise stimu- lated for shared motives. Teacher training must take into account, especially for the teaching of grammar and its relation to writing, that in this framework the interac- tion in the classroom is not simply desirable technique but rather operates as a true development and learning engine. There is a general consensus on the importance of using the language (oral and written) as a tool for learning in any area of the curriculum. However, reflective prac- tices in Language Arts to learn to write and to learn grammar (i.e., metalinguistic activity) are still rare among teachers. We consider that a possible reason for this would be that they are placed in a framework dominated by the dichotomies de- scribed in Table 5. The results of the interview suggest, although only tangentially, the need to rethink the classroom methodologies and, above all, to rethink teacher training in language and linguistics, and especially grammar, which is also based on dialogue and reflection, and not only on prescription. REFERENCES Abad, V. (2015). Los modelos de género discursivo en la planificación y revisión textual: Análisis del uso de modelos en alumnos de 3º ESO [The models of discursive genre in text planning and revision: Analysis of the use of models by late secondary graders]. Unpublished doctoral dissertation. València, Spain: Universitat de València. Andrews, R., Torgerson, C., Beverton, S., Freeman, A., Locke, T., Low, G., Robinson, A. & Zhu, D. (2004). The effect of grammar teaching (sentence combining) in English on 5 to 16-year olds’ accuracy and quality in written composition: Review summary. York, UK: University of York. Andrews, R. (2010). Teaching sentence-level grammar for writing: The evidence so far. In T. Locke (Ed.), Beyond the grammar wars: A resource for teachers and students on developing language knowledge in the English/Literacy classroom (pp. 91–108). New York, NY: Routledge. Barth, B.-M. (1987). L’apprentissage de l’abstraction [The learning of abstraction]. Paris, France: Retz. Bastons, N., Comajoan-Colomé, L., Guasch, O. & Ribas, T. (2017). Les creences del professorat sobre l’ensenyament de la gramàtica a l’aula: cap a una tipologia de docents [Teachers’ beliefs on grammar teaching in primary, secondary and adult education]. Caplletra, 63, 139-64. https://doi.org/10.7203/caplletra.63.10397 Bernárdez, E. (2008). El lenguaje como cultura [Language as culture]. Madrid, Spain: Alianza. Bernárdez, E. (2017). Non-Western concepts of language: Are they of any use for the scientific study of language? Complutense Journal of English Studies, 25, 9-25. http://revistas.ucm.es/index.php/CJES/article/view/58147 Bosque, I. & Gallego, Á.J. (2016). La aplicación de la gramática en el aula: Recursos didácticos clásicos y modernos para la enseñanza de la gramática [Grammar application in the classroom: old and new teaching resources to teach grammar]. Revista de Lingüística Teórica y Aplicada, 54(2), 63-83. Bronckart, J.-P. (2008). Actividad lingüística y construcción de conocimientos [Linguistic activity and the construction of knowledge]. Lectura y Vida: Revista Latinoamericana de Lectura, 29(2), 6–19. Camps, A. (1986). La gramàtica a l’escola bàsica entre els 5 i els 10 anys [Grammar at basic school among 5- to 10-year-old pupils]. Barcelona, Spain: Barcanova. Camps, A. (1998). L’ensenyament de la gramàtica [The teaching of grammar]. In A. Camps & T. Colomer (Eds.), L’ensenyament i l’aprenentatge de la llengua i la literatura en l'educació secundària [Teaching and learning language and literature in secondary education] (pp. 105-126). Barcelona, Spain: Horsori. Camps, A. (2000). Aprendre gramàtica [Learning grammar]. In A. Camps & M. Ferrer (Eds.), Gramàtica a l’aula [Grammar in the classroom] (pp. 101-118). Barcelona, Spain: Graó. TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 29 Camps, A. (2014). Metalinguistic activity in language learning. In T. Ribas, X. Fontich & O. Guasch (Eds.), Grammar at school: Research on metalinguistic activity in language education (pp. 23-40). Brussels, Belgium: Peter Lang. Camps, A. & Ferrer, M. (Eds.) (2000). Gramàtica a l’aula [Grammar in the classroom]. Barcelona, Spain: Graó. Camps, A., & Milian, M. (Eds.) (2000). Metalinguistic activity in learning to write. Amsterdam, The Neth- erlands: Amsterdam University Press. Camps, A., Guasch, O., Milian, M., & Ribas, T. (2000). Metalinguistic activity: The link between writing and learning to write. In A. Camps & M. Milian (Eds.), Metalinguistic activity in learning to write (pp. 103- 124). Amsterdam, The Netherlands: Amsterdam University Press. Carter, R. (ed.) (1990). Knowledge about language and the curriculum: The LINC reader. London, UK: Hod- der and Stoughton. Casas, M. (2014). When does this happen? Recognition of the retrospective usage of the present tense among primary school students. In T. Ribas, X. Fontich & O. Guasch (Eds.), Grammar at school: Re- search on metalinguistic activity in language education (pp. 203-226). Brussels, Belgium: Peter Lang. Castellà, J.M. (1994). ¿Qué gramática para la escuela? Sobre árboles, gramáticas y otras formas de andarse por las ramas [Which grammar for schooling? On parsing, grammars, and other ways to mess around]. Textos de Didáctica de la Lengua y de la Literatura, 2, 15-24. Chartrand, S.-G. (Dir.) (1996). Pour un nouvel enseignement de la grammaire: propositions didactiques [Towards a new teaching of grammar: pedagogic proposals]. Montréal, Canada: Les Éditions Logiques. Connors, R.J. (2000). The erasure of the sentence. College Composition and Communication, 52(1), 96- 128. https://doi.org/10.2307/358546 Coronas, R. (2014). Evaluative analysis of the instructional approach to studying syntactic coordination in ten secondary school textbooks. In T. Ribas, X. Fontich & O. Guasch (Eds.), Grammar at school: Re- search on metalinguistic activity in language education (pp. 75-110). Brussels, Belgium: Peter Lang. Cubero, R., Cubero, M., Santamaría, A., Ignacio, M. J., Prados, M. & De la Mata-Benítez, M. (2008). La educación a través de su discurso: Prácticas educativas y construcción discursiva del conocimiento en el aula [Education via the educational discourse: Educational practices and discursive building of knowledge in the classroom]. Revista de Educación, 346, 71-104. Cuenca, M. J. (1992). Teories gramaticals i ensenyament de llengües [Grammar theories and language teaching]. València, Spain: Tàndem. Culioli, A. (1990). Pour une linguistique de l’énonciation. Vol. I [Towards a linguistics of enunciation. Vol. I]. Paris, France: Ophrys. DeKeiser, R.M. (1998). Beyond focus on form: Cognitive perspectives on learning and practicing second Language grammar. In C. Doughty & J. Williams (Eds.), Focus on form in classroom second language acquisition (pp.42-63). Cambridge, UK: Cambridge University Press. Dolz, J. & Simard, C. (Eds.) (2009). Pratiques d’enseignement grammatical : Points de vue de l’enseignant et de l’élève [Practices of grammar instruction: teacher and student perspectives]. Québec, Canada: PUL. Ellis, R. (1990). Instructed second language acquisition. Oxford, UK: Blackwell. Ellis, R. (2006). Current issues in the teaching of grammar: An SLA perspective. TESOL Quarterly, 40(1), 84-107. https://doi.org/10.2307/40264512 Ellis, R. (2010). Explicit form-focused instruction and second language acquisition. In B. Spolsky & F.M. Hult (Eds.), The handbook of educational linguistics (pp. 437–455). Oxford, UK: Wiley-Blackwell. Engeström, Y. (2011). From design experiments to formative interventions. Theory and Psychology, 21(5), 598-628. https://doi.org/10.1177/0959354311419252 Fisher, C. (2004). La place des représentations des apprenants en didactique de la grammaire [The role of learners’ representations in grammar pedagogy]. In C. Vargas (Dir.), Langue et études de la langue: Approches linguistiques et didactiques [Language and language studies: Linguistic and pedagogical approaches] (pp. 383-393). Aix-en-Provence, France: Université de Provence. Fontich, X. (2014). Grammar and language reflection at school: Checking out the whats and the hows of grammar instruction. In T. Ribas, X. Fontich & O. Guasch (Eds.), Grammar at school: Research on met- alinguistic activity in language education (pp. 255-284). Brussels, Belgium: Peter Lang. 30 A. CAMPS & X. FONTICH Fontich, X. (2016). L1 grammar Instruction and writing: Metalinguistic activity as a teaching and research focus. Language and Linguistics Compass, 10(5), 238-254. https://doi.org/10.1111/lnc3.12184 Fontich, X. (2017). The grammar-writing interplay: Exploring teachers’ concepts for educational change. Communication presented at the SIG Writing Symposium in the 17th EARLI Conference “Education in the crossroads of economy and politics: Role of research in the advancement of public good”, Tam- pere, August 29th - September 2nd. Fontich, X. & Camps, A. (2014). Towards a rationale for research into grammar teaching in schools. Re- search Papers in Education, 29(5), 598-625. https://doi.org/10.1080/02671522.2013.813579 Fontich, X. & Birello, M. (2015). Tendencias de profesorado para el desarrollo de un modelo de enseñanza de la gramática y la escritura [Teachers’ tendencies for the development of a model of grammar and writing instruction]. In L.-P. Cancelas (Ed.), Tendencias en educación lingüística [Tendencies in lan- guage education] (pp. 103-120). Granada, Spain: GEU. Fontich, X. & Camps, A. (2015). Gramática y escritura en la educación secundaria: Estudio de caso sobre los conceptos de los profesores [Grammar and writing in secondary education: A case study on teachers’ concepts]. Tejuelo, 22, 11-27. Fontich, X. & García-Folgado, M.J. (2018). Grammar instruction in the Hispanic area: The case of Spain with attention to empirical studies on metalinguistic activity. L1-Educational Studies in Language and Literature, 18, 1-39. https://doi.org/10.17239/L1ESLL-2018.18.04.02 Garton, A.F. & Pratt, C. (1989). Learning to be literate. Oxford, UK: Blackwell. Gil, R. & Bigas, M. (2014). The use of metalinguistic terms in writing activities in early primary school class- rooms. In T. Ribas, X. Fontich & O. Guasch (Eds.), Grammar at school: Research on metalinguistic activity in language education (pp.141-172). Brussels, Belgium: Peter Lang. Gil, R. (2017). “Mi gato se ha morido”: Análisis de las argumentaciones de futuros maestros ante producciones verbales infantiles erróneas [“Mi gato se ha morido”: Analysis of future teachers’ argumentations facing children erroneous verbal productions]. In A. Camps & T. Ribas (Eds.), El verbo y su enseñanza: Hacia un modelo de enseñanza de la gramática en la actividad reflexiva [The verb and its teaching: Towards a model for grammar teaching based on reflective practice] (pp. 121-132). Madrid, Spain: Octaedro. Gombert, J.-É. (1990). Le développement métalinguistique [Metalinguistic development]. Paris, France: PUF. González-Nieto, L. (2001). Teoría lingüística y enseñanza de la lengua (lingüística para profesores) [Lin- guistic theory and the teaching of language (Linguistics for teachers)]. Madrid, Spain: Cátedra. González-Nieto, L. (2013). Teorías lingüísticas y enseñanza de la Lengua [Linguistic theories and the teach- ing of language]. Textos de Didáctica de la Lengua y de la Literatura, 63, 12-21 Graham, S. & Perin, D. (2007). A meta-analysis of writing instruction for adolescent students. Journal of Educational Psychology, 99(3), 445-476. https://doi.org/10.1037/0022-0663.99.3.445 Grenfell, M. (2000). Modern languages: Beyond Nuffield, and into the 21st century. Language Learning Journal, 22, 23-30. https://doi.org/10.1080/09571730085200211 Guasch, O. & Ribas, T. (2013). La entrevista en la investigación cualitativa en la didáctica de la lengua [The interview in qualitative research in language pedagogy]. Cultura y Educación, 25(4), 483-488. https://doi.org/10.1174/113564013808906852 Harris, R. & Taylor, T.J. (1989). Landmarks in linguistic thought: The Western Tradition from Socrates to Saussure. London, UK: Routledge. Hernández-Navarro, F. & Castelló, M. (2014). Análisis de los géneros de escritura académica, de acuerdo con la percepción de los estudiantes universitarios [Analysis of genres of academic writing according to university students’ perception]. Innovación Educativa, 14(65), 61-80. Hopper, P.J. (1998). Emergent grammar. In M. Tomasello (Ed.), The new psychology of language (pp. 155- 75). Mahwah, NJ: Lawrence Erlbaum. Howatt, A.P.R. & Smith, R. (2014). The history of teaching English as a foreign language, from a British and European perspective. Language and History, 57(1), 75–95. https://doi.org/10.1179/1759753614Z.00000000028 Hudson, R. (2001). Grammar teaching and writing skills: the research evidence. Syntax in the Schools, 17(1), 1-6. TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 31 Hudson, R. (2007). How linguistics has influenced schools in England. Language and Linguistic Compass, 1(4), 227-242. https://doi.org/10.1111/j.1749-818x.2007.00015.x Hudson, R. (2016). Grammar instruction. In C.A. MacArthur, S. Graham & J. Fitzgerald (Eds.), Handbook of writing research (2nd edition) (pp. 288-300). New York, NY: The Guilford Press. Janks, H. (2010). Language as a system of meaning potential: The reading and design of verbal texts. In T. Locke (Ed.), Beyond the grammar wars: A resource for teachers and students on developing language knowledge in the English/Literacy classroom (pp. 151–169). New York, NY: Routledge. Jones, P.T. & Derewianka, B.M. (2016). A brief history of genre pedagogy in Australian curriculum and practice. Sprogforum, 63, 1-16. https://unipress.dk/media/14675/sprogforum_63_jones_derewi- anka.pdf Karmiloff-Smith, A. (1992). The child as linguist. In Karmiloff-Smith, A., Beyond modularity: a developmen- tal perspective on cognitive science (pp. 31-63). Cambridge, MA: MIT Press/Bradford Books. Lantolf, J.P. & M.E. Poehner (2014). Sociocultural theory and the pedagogical imperative in L2 education: Vygotskyan praxis and the research/practice divide. London, UK: Routledge. https://doi.org/10.4324/9780203813850 Lentin, L. (1974). Apprendre à parler à l’enfant de moins de 6 ans [Teaching young infants how to talk]. Repères pour la rénovation de l'enseignement du français à l'école élémentaire, 23, 120-123. www.persee.fr/doc/reper_0755-7906_1974_num_23_1_1232 Llisterri, J. (1996). Preliminary recommendations on spoken texts: EAGLES Documents (Expert Advisory Group on Language Engineering Standards). Bellaterra, Spain: UAB. Retrieved in January 2018 from http://www.ilc.cnr.it/EAGLES96/spokentx/spokentx.html Locke, T. (2009). Grammar and writing: The international debate. In R. Beard, D.A. Myhill, M. Nystrand & J. Riley (Eds.), The SAGE handbook of writing development (pp. 182-193). London, UK: SAGE. https://doi.org/10.4135/9780857021069.n13 Locke, T. (Ed.) (2010). Beyond the grammar wars: A resource for teachers and students on developing language knowledge in the English/Literacy classroom. New York, NY: Routledge. https://doi.org/10.4324/9780203854358 MacArthur, C. A. (2013). Best practices in teaching evaluation and revision. In S. Graham, C. A. MacArthur & J. Fitzgerald (Eds.), Best practices in writing instruction (2nd ed.). (pp. 215-237). New York, NY: Guil- ford. Milian, M. (2005). Reformulation. L1–Educational Studies in Language and Literature, 5(3), 335-351. https://doi.org/10.1007/s10674-005-8560-9 Milian, M. (2014). Working on grammar at school. In T. Ribas, X. Fontich & O. Guasch (Eds.), Grammar at school: Research on metalinguistic activity in language education (pp. 43-74). Brussels, Belgium: Peter Lang. Miller, R. (2011). Vygotsky in perspective. New York, NY: Cambridge University Press. https://doi.org/10.1017/CBO9780511736582 Myhill, D. (2000). Misconceptions and difficulties in the acquisition of metalinguistic knowledge. Language and Education, 14(3), 151–163. https://doi.org/10.1080/09500780008666787. Myhill, D., Jones, S. M., Lines, H. & Watson, A. (2012). Re-thinking grammar: The impact of embedded grammar teaching on students’ writing and students’ metalinguistic understanding. Research Papers in Education, 27(2), 139–166. https://doi.org/10.1080/02671522.2011.637640 Myhill, D.A. & Jones, S.M. (2015). Conceptualising metalinguistic understanding in writing. Culture and Education, 27(4), 839–867. https://doi.org/10.1080/11356405.2015.1089387 Myhill, D., Jones, S.M. & Wilson, A. (2016). Writing conversations: fostering metalinguistic discussion about writing. Research Papers in Education, 31(1), 23-44. https://doi.org/10.1080/02671522.2016.1106694 Nadeau, M. & Fisher, C. (2011). Les connaissances implicites et explicites en grammaire: quelle importance pour l’enseignement? Quelles conséquences? [Implicit and explicit knowledge in grammar: Which relevance for teaching? Which consequences?]. Bellaterra Journal of Teaching and Learning Lan- guage and Literature, 4(4), 1–31. Nordlinger, R. & Bresnan, J. (2011). Lexical-Functional Grammar: Interactions between morphology and syntax. In R.D. Borsley & K. Börjars (Eds.), Non-transformational syntax: Formal and explicit models of grammar (pp.112-140). Hoboken, NJ: Blackwell. 32 A. CAMPS & X. FONTICH Nupponen, A.-M., Jeskanen, S. & Rättyä, K. (2019). Recognizing linguistic concepts related to sentence structures―student teachers reflecting L1 and L2 textbooks. L1-Educational Studies in Language and Literature, 19, 1-25. https://doi.org/10.17239/L1ESLL-2019.19.02.04 Peytard, J. & Genouvrier, É. (1970). Linguistique et enseignement du français [Linguistics and the teaching of French]. Paris, France: Larousse. Rättyä, K. (2013). Languaging and visualisation method for grammar teaching: A conceptual change theory perspective. English Teaching: Practice and Critique, 12(3), 87-101. Ribas, T., Fontich, X. & Guasch, O. (Eds.) (2014). Grammar at school: Research on metalinguistic activity in language education. Brussels, Belgium: Peter Lang. https://doi.org/10.3726/978-3-0352-6490-6 Rodríguez-Gonzalo, C. (2012). La enseñanza de la gramática: las relaciones entre la reflexión y el uso lingüístico [The teaching of grammar: the interplay between reflection and language use]. Revista Iberoamericana de Educación, 59, 87-118. Rodríguez-Gonzalo, C. (2015). A classroom intervention on Spanish grammar and writing: the use of the past tense with secondary school students. Culture and Education, 27(4), 879-898, https://doi.org/10.1080/11356405.2015.1089388 Roth, F.P., Speece, D.L., Cooper, D.H. & de la Paz, S. (1996). Unresolved mysteries: How do metalinguistic and narrative skills connect with early reading? The Journal of Special Education, 30(3), 257–277. https://doi.org/10.1177/002246699603000303 Schicker, J. (2018). Teaching writing and grammar through critical thinking in a genre-based upper-level undergraduate German class. Die Unterrichtspraxis/Teaching German, 51(1), 63-75. Schwarz, B., Dreyfus, T. & Hershkowitz, R. (Eds.) (2009). Transformation of knowledge through classroom interaction. London, UK: Routledge. Serra, M., Serrat, E., Solé, R., Bel, A. & Aparici, M. (2000). La adquisición del lenguaje [Language acquisi- tion]. Barcelona, Spain: Ariel. Simard, C., Dufays, J.-L., Dolz, J. & Garcia-Debanc, C. (2010). Didactique du français langue première [Pe- dagogy of French L1]. Brussels, Belgium: De Boeck. https://doi.org/10.3917/dbu.simar.2010.01 Taylor, T.J. (1997). Theorizing language: Analysis, normativity, rhetoric, history. Oxford, UK: Pergamon Press. Taylor, T.J. (2000). Language constructing language: The implications of reflexivity for linguistic theory. Language Sciences, 22(4), 83-99. Tisset, C. (2010). Enseigner la langue française à l’école. La grammaire, l’orthographe et la conjugaison [Teaching French language at school: Grammar, orthography, and conjugation]. Paris, France: Ha- chette. Trévise, A. (2009). Les textes officiels français pour les langues vivantes et l’enseignement/apprentissage de la grammaire: L’exemple de la détermination nominale en anglais [French official texts for modern languages and the teaching-learning of grammar: The example of noun determination in English]. In M.-L. Elalouf (Ed.), Diptyque 16 (pp. 103-124). Namûr, France: Presses Universitaires de Namûr. Tuson, J. (1980). Teorías gramaticales y análisis sintáctico [Grammar theories and syntactic analysis]. Bar- celona, Spain: Teide. Van Rijt, J. & Coppen, P.-A. (2017). Bridging the gap between linguistic theory and L1 grammar education: Experts’ views on essential linguistic concepts. Language Awareness, 26(4), 360-380. https://doi.org/10.1080/09658416.2017.1410552 Van Rijt, J., Wijnands, A. & Coppen, P.-A. (2019). Dutch teachers’ beliefs on linguistic concepts and reflec- tive judgement in grammar teaching. L1-Educational Studies in Language and Literature, in this Spe- cial Issue. Vygotsky, L. S. (1987). Thinking and speech. In R. Rieber & A. Carton (Eds.), The collected works of Lev Vygotsky. N. Minick, Trans. Vol. 1 (pp. V + 285). New York, NY: Plenum. Watson, A. (2015a). Conceptualisations of grammar teaching: L1 English teachers’ beliefs about teaching grammar for writing. Language Awareness, 24(1), 1-14. https://doi.org/10.1080/09658416.2013.828736 Watson, A. (2015b). The problem of grammar teaching: a case study of the relationship between a teach- er's beliefs and pedagogical practice. Language and Education, 29(4), 332-346. https://doi.org/10.1080/09500782.2015.1016955 TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 33 Wyse, D. (2001). Grammar for writing? A critical review of empirical evidence. British Journal of Educa- tional Studies, 49(4), 411-427. https://doi.org/10.1111/1467-8527.t01-1-00185 Zayas, F. (2012). Los géneros discursivos y la enseñanza de la composición escrita [Discursive genres and the teaching of writing composition]. Revista Iberoamericana de Educación, 59, 63-85. APPENDIX – SAMPLING PROCESS FOR THE LIKERT QUESTIONNAIRE The Likert questionnaire was conducted in 2012. Schools selected are located in the metropolitan area of Barcelona, in the district of Vallès Occidental (population: 900,000 inhabitants; immigration rate: around 10%; density: 1,500/km 2; average density in the Catalan Autonomous Community: 234/km 2). The established objective was to obtain above 100 completed questionnaires in total (comprising Primary and Secondary teachers); the final number of questionnaires received was 94. The frame- work of the sample was the total number of state schools in the area: 40 at Primary and 58 at Secondary level. The selection was made following a stratification proce- dure according to immigration rate (see Tables 1 and 2). The data from the question- naire were processed using the SPSS program; for analysis we relied on the expertise of Dr Anna Cuxart and Dr David Roche, from the UPF (Universitat Pompeu Fabra) in Barcelona. The sampling of Primary schools is shown in Table A1. Once the 13 centres were selected, their principals sent a list with the language teachers’ email addresses (87). The total number of teachers taking part (44) represented a 50,1% of the total of teachers contacted. Table 2 shows the sampling of Secondary schools. It was decided to take 11 centres that work regularly in collaboration with the UAB (Universitat Autònoma de Barcelona) and to expand the sample with a further nine centres. Out of the 20 centres selected, 11 were keen to take part in the questionnaire (making no differentiation between Subject Catalan and Subject Spanish). The total of ques- tionnaires received numbered 50 (representing 59,5% of the teachers contacted). 34 A. CAMPS & X. FONTICH Table A1: Universe, sampling according to immigration rate, and number of schools selected in primary Universe Total Stratification according to immigration rate 0% < 5% be- be- >15% CAEP* tween tween 5,1- 10,1- 10% 15% Total schools in 162 31 98 19 11 3 12 Vallès Occidental Distribution in % 100% 19,1% 60,5% 11,7% 6,8% 1,9% 7,4% Total schools for 40 4 27 6 3 0 1 practice in UAB Distribution in % 100% 10,0% 67,5% 15,0% 7,5% 0,0% 2,5% Total schools of 13 2 8 2 1 0 1 the sample Distribution in % 100% 15,4% 61,5% 15,4% 7,7% 0,0% 7,7% *CAEP: Centre d’Atenció Especial (Centre for Special Attention) Table A2: Universe, sampling according to immigration rate, and number of schools selected in Secondary Universe Total Stratification according to immigration rate 0% < 5% between Between >15% CAEP* 5,1- 10% 10,1-15% Total schools 58 4 34 15 3 2 6 in Vallès Occi- dental Distribution 100% 6,9% 58,6% 25,9% 5,2% 3,4% 10,3% in % Total schools for 11 0 10 1 0 0 1 practice in UAB Distribution 100% 0,0% 90,9% 9,1% 0,0% 0,0% 9,1% in % Total schools 20 1 10+2 0+5 1 1 1 of the sample Distribution 100% 5,0% 60,0% 25,0% 5,0% 5,0% 5% in % *CAEP: Centre d’Atenció Especial (Centre of Special Attention) TEACHERS’ CONCEPTS ON TEACHING GRAMMAR 35 Socio-demographic profiles of the participants are shown in Tables A3 and A4, while the modality of instruction can be seen in Table A5 below. Table A6 shows (for Sec- ondary teachers only) whether they teach in Subject Catalan or Subject Spanish, and Table A7 shows the years of service. Table A3: Gender Frequency Percent Male 18 19,1 Female 76 80,9 Total 94 100,0 Table A4: Degree Frequency Percent a) Only 3-year Primary Degree* 36 38,3 b) Only 5-year Degree** 43 45,7 Both a) and b) 6 6,4 Other degrees (master, doctorate) 9 9,6 Total 94 100,0 *This entitles to teach at Primary **This entitles to teach at Secondary and is language specific (Spanish, Catalan, English, etc.) Table A5: Modality of instruction (Primary, Secondary, both) Modality of instruc- Submodality Number of teachers per mo- tion dality and submodality Primary only 43 Secondary only (total) 50 Lower sec. 24 Lower and Upper 23 sec. Upper sec. 3 Primary and Upper 1 sec. Total 94 36 A. CAMPS & X. FONTICH Table A6: Teachers at Subject Catalan or Subject Spanish Frequency Percent Subject Catalan 25 49,0 Subject Spanish 25 49,0 Subject Catalan and Spanish 1 2,0 Total 51 100,0 Table A7: Years of service Frequency Percent 0 to 5 years 14 14,9 6 to 10 years 13 13,8 11 to 24 years 34 36,2 25 years or more 33 35,1 Total 94 100,0