Idioms and fixed expressions - Implications of cognitive linguistics for the selection of web-based teaching exercises

Table of contents

1 Introduction

2 Theoretical approaches
2.1 Transfer in idiom learning
2.2 Cognitive semantics and the arrangement of idiom dictionaries
2.3 The scale of idiomaticity and the role of (conceptual) metaphors
2.4 Teaching implications for collocations

3 Analysis of current e-learning websites for idioms
3.1 http://english-zone.com
3.2 http://esl.about.com
3.3 www.idiomconnection.com
3.4 http://www.readwritethink.org/materials/idioms
3.5 http://a4esl.org/q/h/idioms.html
3.6 www.akademie.de/fuehrung-organisation/management/kurse/englische- korrespondenz/erfolgskontrolle/idioms-test.html

4 Summary and outlook

5 Conclusion

6 References

1 Introduction

Idioms are an important factor in native-like discourse by an English as a foreign language (EFL) speaker. According to Fernando (1996) “the sheer number of idioms and their high frequency in discourse make them an important aspect of vocabulary acquisition and language learning in general”. Many linguists claim that idioms require special attention in language programs and should not be relegated to a position of secondary importance in the curriculum.

However, classroom learning of EFL is - at least in Germany - not very concerned with the teaching of idioms and fixed expressions, as analyses from current and historical textbooks have shown. This results in the necessity for the teacher to make up own exercises or series of lessons to put the teaching of fixed expressions into the curriculum. The cognitive approach to language learning provides useful aspects and implications on how to organize idiom learning in a classroom context.

The aim of our paper is to use the cognitive linguistics approach towards idioms and other fixed expressions in English to analyse different homepages on the World Wide Web. In order to do this we will first present a short survey of different cognitive approaches to idioms and their didactic implications for teaching exercises. We will then go on with an analysis of six websites with respect to how the content, structure and exercises follow the implications given by the presented cognitive approaches.

After a discussion of the positive and negative aspects we found on the web, we try to give a proposition on how an idiom learning website could be structured and how exercises could look like, when they try to follow a cognitive approach.

2 Theoretical approaches

The following chapter will present some of the basic ideas of cognitive linguists that can be accounted for when looking for parameters that should be incorporated in learning websites. The central findings of these authors, as they provide the basis for the evaluation of different websites, are highlighted by being printed in an indented format.

2.1 Transfer in idiom learning

Irujo (1986a) conducted a study which was focused on the question how transfer influences the acquisition of idioms in English for foreign language learners. Her study was based on the presumption that Second Language Learners have difficulties in using idioms. These difficulties are supposed to be the result of transfer problems, since previous learning in her opinion always effects subsequent learning. Irujo’s approach follows a rather “traditional view” of idioms because she defines them as “a conventionalized expression whose meaning cannot be determined from the meaning of its parts” (ibid, 288). Conventionalized in this case means that a native speaker immediately understands what has been said and does not have to analyse and interpret the meaning as in a metaphor.

Although in the psycholinguistic tradition there is not much interest on the idea of transfer it still provides some useful hypotheses for research on idiom learning. This is the case when authors see language learning as “creative construction” rather than imitating habits. One can also see transfer as a part of this creative construction, when the transferred parts are not seen as habits but rather as domains of language (Gass & Selinker 1983).

The strong version of the contrastive analysis hypothesis proposed that errors made because of wrong transfer of language domains can be predicted. Irujo rather follows the so called “weak version” of this hypotheses, which explains that errors and difficulties occur more frequently when the target language and first language differ only slightly.

Based on these theoretical background she formulates three hypotheses guiding her study:

- identical idioms show evidence of positive transfer and are easiest to comprehend
- similar idioms show evidence of negative transfer, especially when it comes to producing idioms
- different idioms show no signs of either positive or negative transfer (ibid., 290).

The results of her study subjects, who consisted of twelve native speakers of Spanish from Venezuela that were also proficient speakers of English (avg. TOEFL score was 570), supported all three hypotheses. They understood identical idioms as well as similar ones better than completely different idioms, while producing more identical idioms correctly than similar or different ones. The comprehension of similar idioms was on a high level although the production showed interference from Spanish. And finally, the production of different idioms did not show interference from the first language.

Although she admits that there is a need for further empirical studies on the topic, Irujo suggests to apply her findings. Based on the discussion of the results of her study she suggests to adopt certain principles in teaching idioms in a classroom context. Some of which might also help in analysing web-based teaching material. The most important consequence of her study is the call to make use of the knowledge of idioms in the first language in order to use positive transfer and avoid negative interference. When we consider this parameter, the findings of our web analysis can only show the point of view of a teacher in English classes that have a mainly native German speaking background.

Irujo also asks for avoiding the teaching of “infrequent, highly colloquial idioms with difficult vocabulary” (Irujo 1986a, 298), because production of those idioms by non-native speakers would be very hard and mostly incorrect. Here, the term appropriateness plays an important role. It is hard for a non-native speakers to identify certain situations or settings where those idioms are applicable and where not (ibid, 299). Another claim Irujo makes is that activities should be used which encourage not only the recognition and understanding of idioms but also the production. This does not only account for high-level classes but also for beginners and intermediates and will be one of the more important parameters in the analysis in chapter 3.

2.2 Cognitive semantics and the arrangement of idiom dictionaries

Kövecses (2001) gives a detailed account of the usability of conceptual metaphors in foreign language teaching (FLT). He points out that the central theme underlying cognitive linguistics is the concept of “motivated meaning”. This concept is shown to be useful in the context of learning and teaching idioms. Kövecses focuses on five different areas on where the cognitivistic approach can provide answers. However, for the purpose of this paper we would like to focus on two things. Firstly we concentrate on his findings of what the most common idioms are and secondly on his account of an “ideal” arrangement for idiom dictionaries, because it could give us hints also for the arrangement of an e-learning website.

Based on the cognitive approach he predicts that those are the most common idioms that have a source, that can be experienced most directly (ibid., 88), which therefore should be the human body, especially the hand. This prediction was proven by some empiric studies that e.g. counted body metaphors in an idiom dictionary. One of these has shown that about one sixth of all the listed idioms were related to the human body (ibid, 89). He therefore claims that idioms that have to do with the human body should be in the foreground of FLT.

The conceptual organization related to the processing of idioms should provide the basis for an arrangement of idioms. This means, that the source as well as the target domain of an idiom and its corresponding conceptual metaphor should be incorporated in a system, by which dictionaries (and for our case: learner websites) should be arranged.

Neither an alphabetical listing nor a key-word based arrangement can fulfil such a claim (ibid., 90f.). A thesaurus-like listing provides a first step, since idioms are sorted under a word or phrase that resembles the content or the expression stated in the particular idioms. One could therefore say that such an arrangement provides only the target domain of a conceptual structure. His proposal of an ideal idiom arrangement provides the learner with a “metaphor-based arrangement” (ibid., 93) in which the source as well as the target domain are beingexplicitly included (ibid., 94). This arrangement could look like this:

illustration not visible in this excerpt

Figure 1 (Kövecses 2001, 95)

The problem is, that each source domain (in this case FIRE) can obviously provide conceptual metaphors for a variety of different targets, such as LOVE, IMAGINATION, CONFLICT, ENERGY or ENTHUSIASM. This is called “the scope of metaphor” (Kövecses, 2000) and should also be included into an idiom arrangement:

illustration not visible in this excerpt

Figure2: Several target domains with a single shared source (Kövecses 2001, 95)

For Kövecses it seems important that a dictionary should only provide one single entry for each source domain that connects to all possible target domains. By doing that, one could assure “that certain aspects of abstract concepts could be seen as directly representable as mapped versions of certain concrete source concepts” (Kövecses 2001, 95).

2.3 The scale of idiomaticity and the role of (conceptual) metaphors

Lennon (1998) describes an idiomatic scale in the language production of second language learners of English. On this scale “non-idiomatic discourse is non-native-like rather than erroneous” (ibid, 11). This means that the discourse is correct on a level referring to morphemes, words, etc.^[1] but is graded as not acceptable by native speakers as part of their language.

The use of idioms is so important to native speakers that sometimes faulty idioms are seen as correct although they “defy grammatical conventions” (ibid., 12). This does not mean that native English is made up only from idiomatic language. It rather requires a “blend of formulaic conventionalised language and language generated for the nonce” (ibid. 14) in order to arrive on native-like level. This statement from Fillmore (1979) is revisited by Lennon when he argues for “collocational and combinational co-occurrence probabilities” (ibid., 15) that define whether discourse is conventionalised to a rather high or low degree. Another problem Lennon addresses is appropriateness. He points out the importance of the situational context, that can - in some instances - require a high level of conventionalised language, e.g. when offering one’s condolences at a funeral.

The parameters that define, whether discourse is more or less idiomatic (with idioms marking highest degree of idiomaticity) are:

“limited substitutability or productivity

indivisibility (multi-word unit)

semantic opacity “ (ibid., 16).

In that line of argumentation he reasons that idioms are memorised as “more or less pre-fabricated language together with their substitution restrictions and possibilities” (ibid, 16). This can be seen as one of the distinctive features of the mental lexicon towards a dictionary. Given this distinction, it becomes obvious that the teaching of idioms or other multi-word units is of central importance to the teaching of a second language. The complexity of multi-word units ranges from consisting of two collocational nouns (as in “knife and fork”, “salt and pepper”) to collocations including at least one metaphorical element (as in “dead drunk”) and full idioms that require a verb phrase and whose metaphorical meaning spans the whole phrase^[2]. Since Lakoff and Johnson (1980) have shown the importance and effectiveness of the human metaphorical system it becomes clear that native speakers have “internalised the underlying metaphorical nature of (…) language” (Lennon, 1998) to a degree where they need to make a conscious effort in order to carve out the metaphor hidden inside many expressions of everyday speech. This could be done by non-native speakers with a good proficiency even better than by native speakers, which often do not see (or search for) the meaning behind an expression.

This has got pedagogical implications, that become important for the task of this paper. Lennon’s claims are:

exercises should use a problem solving approach

exercises that try to activate the metaphorical analysis of fixed expressions

exercises that provoke discussion and comparison to the L1

make students aware of the similarities of L1 and English.

These implications go back to an approach called “effort after meaning” by Bartlett (1932), which states that in processing unknown expressions or stories containing unfamiliar practices learners tend to memorise these stories in a way that eventually fits their expectations and they try to transfer from their former knowledge.

[...]

^[1] Lennon refers to the different levels of speech in terms of domain and extent, while the first one describes the linguistic and semantic circumstances, that determine if an expression is correct or not, and the latter one is the level that needs to be adjusted in order to get a correct statement (from word to the whole sentence).

^[2] In this case we leave out the labelling of proverbs, stock phrases, social formulae and stereotypical phrases that exemplify an even higher degree of conventionalised language.