Construction-based language learning and teaching

De Knop, Sabine [USL-B]

In the last decades Construction Grammar (CxG) has established itself as a usage-based approach which can compete with functional or projectionistic descriptions of language. It offers an all-encompassing framework which makes it possible to describe linguistic units, so-called ‘constructions’, at different levels of abstraction and of various degrees of complexity, going from morphological units to more complex syntactic structures. But the fruitful potential of CxG is not limited to theoretical studies, it can also be extended to more applied domains of linguistic research like foreign language learning (FLL) and teaching (FLT). Inspired by studies on the acquisition of constructions by natives (see among others Behrens 2009; Goldberg et al. 2004), several studies in FLL have brought evidence about the ontological status of constructions in the learners’ constructicon (see sentence-sorting experiments by Gries & Wulff (2005), Valenzuela Manzanares & Rojo Lopez (2008), but also more recently Baicchi (2016), or De Knop & Mollica (2016)). As a consequence, CxG has been adopted in various applied studies (see among others the collective volume by De Knop & Gilquin (2016) or the special issue by Ellis & Cadierno (2009)) as a full-fledged model which has several assets to offer for issues in FLL and FLT. The present chapter aims at zooming in more deeply onto these assets and at covering a wide range of issues in the applications of CxG to FLL and FLT. The definition of constructions as entrenched meaningful structures which ally form and function (Goldberg 1995; 2006; 2019) refers to dimensions which can be exploited in FLL. Learning a FL means learning language-specific L2-constructions (Boas & Ziem 2018; Ziem & Boas 2017) which can differ from L1-constructions.  As constructions have a meaning of their own, learners can infer the meaning of new constructional instantiations in L2 from the knowledge of abstract constructions (Römer et al. 2014: 967).  Words - and more specifically verbs - acquire their meaning in the construction (Ellis & Ferreira 2009). This observation allows to understand why some verbs can be used differently from their valency or functional description, e.g. the non-transitive verb dream in the caused motion construction She dreams him into her life.  Constructional units converge with chunks, formulaic sequences, etc. The learning of constructions takes place holistically and in the same way as with chunks (Wee 2007: 25; Ellis 2003; Ellis & Cadierno 2009). It has been shown that chunking is an efficient way of learning structures which do not exist in L1 (Ellis 2009: 147; Handwerker 2008; Handwerker & Madlener 2006). This applies more specifically to the learning of an L2 which belongs to a different typological class than the L1 (cf. Herbst 2016: 43). E.g. for French-speaking learners of German, an applicative construction (see Michaelis & Ruppenhofer 2001) like Der Weihnachtsmann hat die Kinder reichlich beschenkt (Lit. ‘Father Christmas has the children generously offered-presents’) is problematic as the applicative construction cannot be explained by lexical rules focusing on the be-verb.  Constructions have a certain degree of frequency. Therefore, learning L2-constructions means focusing on well-entrenched and usage-based linguistic units. Each language has an inventory of constructions which are organized in a network, the so-called constructicon. It includes more abstract constructions, e.g. the caused motion or the applicative construction, but also concrete instantiations of abstract constructions. Moreover, because CxG does not separate between grammar and the lexicon or phraseology but advocates a lexicon-grammar continuum, idioms are also considered as specific instantiations of (more abstract) constructions. In the constructicon constructions are semantically related to each other by inheritance links. Ruiz de Mendoza & Agustin Llasch (2016) have demonstrated that the caused motion construction and the resultative construction are closely linked to each other (see also Goldberg (1995) and Goldberg & Jackendoff’s (2004) study on resultatives). In the same way, German pleonastic constructions like e.g. Die Mutter setzt das Kind auf das Pferd drauf (Lit. ‘The mother sets the child on the horse upon’) are related to motion constructions with directional adverbs like Das Kind läuft bis auf den Berg hinauf (Lit. ‘The child runs up to the mountain upon’). A network approach allows to describe constructions simultaneously and to explore how the different constructional slots can be filled semantically, e.g. which verb can be used in which construction. At the same time, it can be shown how one and the same verb can appear in different constructions. These are logico-economical arguments in favor of CxG (Stefanowitsch (2011: 182)). The chapter further deals with teaching issues, it illustrates how CxG can facilitate the teaching of L2-constructions (Vujic 2016). First of all, learners have to be made aware of differences between L1 and L2 (Ellis et al 2016: 300), e.g. the preference for verbless directives in the satellite-framed language German as opposed to verb-framed constructions in French (Talmy 2000). Accordingly, typological aspects should be integrated into the constructionist approach. Some more specific teaching strategies can foster the learning of the L2-constructions.  One possibility is the ‘scaffolding strategy’ (Klewitz 2017). It is advisable to start from simple constructions and let more complex and related constructions be discovered (see also Holme (2010) and Römer et al. (2014: 967)). This can be illustrated with pleonastic constructions, like Die Mutter setzt das Kind auf das Pferd drauf (Lit. ‘The mother sets the child on the horse upon’) as an extension of simple motion constructions like Die Mutter setzt das Kind auf das Pferd (Lit. ‘The mother sets the child on the horse’). This implies a “more active involvement of students in uncovering properties of constructions that they already use, as well as properties that they are expected to acquire” (Wee 2007: 29).  Another teaching strategy is ‘structural priming’ (Gries & Wulff 2009). Psycholinguistic studies (Ellis et al. 2016: 40; Hartsuiker et al. 2004; Loebell & Bock 2003) have shown that when hearing or reading a specific structure in L2, learners tend to integrate and reproduce the same structure with other words, as illustrated by a study with Italian-speaking learners of German verbless directives (see De Knop & Mollica 2021).  Last but not least, since the advent of Cognitive Linguistics, it is well-known that our language is the result of our embodied experience with reality (Lakoff & Johnson 1999). The embodied cognition principle offers fruitful avenues for the teaching of foreign constructions – as has been shown in De Knop (2020) or Skulmowski & Rey (2018). To conclude: By being usage-based and by dispensing with a complicated descriptive apparatus, the constructionist approach allows a simplified view on language teaching (Herbst 2016: 25). Moreover, it can be combined with other research fields like corpus linguistics, language typology and contrastive analysis (Herbst 2016).