How native and non-native speakers of English interpret unfamiliar formulaic sequences

Abstract

This study examines whether native and non-native speakers faced with unfamiliar formulaic expressions use the same tactics for working out what they mean. The test items needed to be semantically opaque, used in an authentic context and unknown to all participants. Ten obsolete expressions were selected from the historical novels of Georgette Heyer. First-language English speakers and UK-resident classroom-taught learners of English as a foreign language were individually presented with the expressions in their original context, and asked to work out what they meant. Analysis of their comments revealed that the native speakers deployed significantly more context and analogy. The non-native speakers were much more likely than native speakers to refer to individual unknown words. Whilst it seems that first-language users take a more holistic approach to linguistic input than classroom-taught second-language learners do, the findings may suggest that learners adopt increasingly ‘nativelike’ strategies as proficiency increases.

Keywords:

• Analogy
• comprehension
• context
• formulaic language
• idioms
• language learning

Notes

1. LOL is defined by the Oxford English Dictionary (OED) and American Heritage Dictionary as meaning ‘laugh(ing) out loud’ but other interpretations, including ‘lots of love’, are also attested, most famously in the usage of UK Prime Minister David Cameron, http://www.businessinsider.com/david-cameron-lol-lots-of-love-rebekah-brooks-2012-5

2. Example taken from http://idioms.thefreedictionary.com/kick+over+the+traces

3. Wray (2008b) also considers the value of delaying language teaching in schools until all class members have exited the implicit learning period, to avoid teaching learners with fundamentally different approaches to learning.

4. If a phrase is unfamiliar to a native speaker, it fails one of the tests normally applied to establish formulaicity (Wray and Namba, 2003; Wray, 2008a): having a known conventionalised meaning. But other tests will confirm it as formulaic.

5. Our choice of the term ‘non-native speaker’ is deliberate, even though in some contexts it underrates the legitimacy of the second-language speaker as an expert user. We are interested in whether the fact of learning a language after infancy and in a classroom impacts on how unknown expressions are interrogated.

6. This task has no formal status in language testing and we cannot by any means take it that increased performance in this task would exactly mirror a wider measure of proficiency. However, we will treat their test performance as a proxy for proficiency for the purposes of this account.

7. The meanings of the expressions are given in Appendix 3.

8. However, when Zuo (2008) compared the strategies used on unknown English idioms with more and less direct equivalents in Chinese, she found no major differences, suggesting that analogy was not playing a major part for her participants.

9. Technically this is not quite true, because a formulaic frame could be populated with novel slot-fillers, e.g. The Emperor penguin pulled the snail’s leg. A truly novel sentence is one that does not deploy a formulaic frame, e.g. Leave seven potatoes and sign the undertaking. Having said that, emergent grammar models hold that any grammatical sentence is based on a frame. But this need not concern us here, because our focus is on the semantics rather than the grammar.

10. The capacity to infer the meanings of unknown words in this way has been demonstrated in dogs. Kaminski, Call and Fischer (2004) report that a border collie, Rico, ‘knew the labels of over 200 different items. He inferred the names of novel items by exclusion learning and correctly retrieved those items right away as well as 4 weeks after the initial exposure’ (1682).