Replication Data for: Metaphors in high-stakes language exams

Abstract

The dataset and R code here provide documentation for the chapter Metaphors in high-stakes language exams. Chapter abstract: Lakoff and Johnson’s Conceptual Metaphor Theory advances the view that metaphor is a fundamental cognitive process defining our understanding of reality: “the essence of metaphor is understanding and experiencing one kind of thing [e.g. love] in terms of another [e.g. a journey]” (Lakoff & Johnson, 1980, p. 5). Such metaphors in thought (conceptual metaphors) are reflected as metaphors in language, i.e. by the words and expressions we produce (linguistic metaphors). Empirical research has since confirmed that linguistic metaphor is ubiquitous in both L1 and L2 language (see e.g. Nacey, 2013; Steen et al., 2010). While metaphor therefore necessarily plays an important role in language learning, some scholars suggest that processing figurative language may pose challenges for L2 speakers of a language, who are less familiar with cultural conventions and connotations, and lack a figurative language repertoire (see e.g. Littlemore & Low, 2006). Pickins notes that this may be particularly true when learners are asked to interpret literary texts, which may “stretch the resources of the language to convey a particular poetic vision”. Metaphor is thus thought to be “difficult” for L2 learners, although as Low pointed out already in 1988 (p. 137) “it would be helpful to know whether the ways in which learners learn to cope with metaphor are similar from person to person”. This paper addresses that acknowledged need by investigating how L2 learners of Norwegian respond to a task requiring them to interpret a literary metaphor and incorporate that metaphor in a text about their own lives: a task involving both receptive and productive metaphorical competence. The empirical data consists of 22 texts collected in the Norwegian Second Language Corpus (ASK), written by L2 Norwegian learners as part of the Test i norsk – hoyere niva [Test in Norwegian – higher level]. This test is a high-stakes language test primarily intended for immigrants to Norway who need to document their language skills for employment or for admission to Norwegian universities and colleges – a situation where students may want to employ low-risk strategies so as not to be penalized for apparent lack of language proficiency (see Littlemore & Low, 2006, p. 70). The texts under investigation – comprising roughly 10,000 words in all – were produced by informants with one of eight different L1s and from one of twelve different countries, the entire set of responses to an identical task. They were instructed to write a text incorporating their own opinions and experiences of friendship with the message(s) in the Kolbjorn Falkeid poem Det er langt mellom venner [It is far between friends]: It is far between friends. / Between friends lie many acquaintances / and much talk. / Friends are like bright, little cottages / far away in the mountain darkness. / You cannot mistake them for anything else. [own translation] At the poem’s core is metaphorical simile steeped in the background of the Norwegian “hyttekultur”, the tradition of enjoying (often) primitive and isolated cabins in the wilderness as a means of temporarily escaping from both the demands of daily life and the temptation of modern conveniences including electricity and running water This study focuses on the degree to which the informants themselves produce metaphor in their response manifesting their understanding of the poem (even though the exam instructions included no explicit mention of metaphor). Following Gibbs (1994, pp. 116-118), ‘understanding’ is here defined as a four-step process of comprehension (creating immediate meaning), recognition (recognizing the metaphor as metaphor), interpretation (analyzing and/or extending the metaphor) and appreciation (aesthetic judgement); Gibbs explains that while comprehension – the grasping of meaning – is crucial, the other three steps are later, and optional, products of understanding. Metaphor identification in the L2 texts is carried out using the Scandinavian version of the Metaphor Identification Procedure Vrije Amsterdam, which requires analysis of each word for metaphorical status (Nacey, Greve, & Falck, 2019). Subsequent analysis focuses upon metaphor density (i.e., how much metaphor is produced), as well as the role of the identified metaphor clusters (i.e., what is the function of metaphor).