Is alliteration mnemonic without awareness-raising?

Abstract

English phraseology abounds with alliterative multiword units (e.g., slippery slope), which suggests that alliterative word strings have a comparative advantage to become stock phrases. One plausible explanation for this advantage is that alliterative word strings are relatively memorable, all else being equal, although there is little directly pertinent empirical evidence. In the context of instructed second language acquisition, it has been suggested that alliteration has a good potential to facilitate the learning of multiword units (MWU), at least when learners are made aware of the presence of alliteration in these units. In the present article, we investigate whether alliteration also makes MWUs comparatively memorable also in the absence of awareness-raising. English alliterative phrases and matched controls were dictated to EFL learners whose ability to recall the phrases was subsequently gauged in unannounced tests. We found a modest, short-lived mnemonic effect of alliteration. This points to an important role for awareness-raising if teachers or materials writers wish to help learners capitalise on the mnemonic advantage afforded by alliteration.

Keywords:

• L2 multiword units
• alliteration
• memory
• recall
• awareness-raising

Acknowledgements

For the statistical analyses we used online calculators generously made available by Richard Lowry (1998–2012). Values of Spearman's rho were calculated using Wessa (2012). Effect sizes and confidence intervals were calculated using the ESCI (Exploratory Software for Confidence Intervals) software accompanying Cumming (2012). For details see http://www.latrobe.edu.au/psy/documents/ESCI-readme.pdf.

Notes

¹Calculating effect sizes, such as Cohen's d, has become an alternative for significance testing in many research disciplines (for a critique of the prioritisation of p-values, see, e.g. Cohen, 1994), and it is becoming increasingly recommended in applied linguistics as well. Effect sizes indicate the magnitude of treatment effects and do not depend on sample size, unlike significance tests. The effect size (d) we report here and elsewhere in this article is the standardised difference between the two means (in this case, the mean scores on alliterative v. non-alliterative sets of phrases). The ‘standardiser’ in the calculation is the pooled inferential standard deviation of the scores for each condition (see Cumming, 2012, for details).

²Lindstromberg and Boers (2008a) did not state effect sizes. The effect sizes given here were calculated on the basis of the data from their study. Cohen (1988) tentatively suggests characterising a d-value between .5 and .8 as ‘medium’ and above .8 as ‘large’. In behavioural and educational studies a d-value as high as 0.88 certainly qualifies as substantive (e.g. Grissom & Kim, 2012, pp. 128–130).

³Boers et al. (2012) did not state an effect size either. We calculated this from their data.

⁴The recall tests used in our experiments required the participants to try and recall phrases that were previously dictated to them. It is safe to say that our participants therefore recalled the target phrases from episodic memory, not semantic memory (Nadel & Hardt, 2011; Tulving, 1972, 1985). Items retrieved from episodic memory are accompanied by recollections about the event in which the item was encountered or used (Tulving, 1993, 2002). In this particular case, that event was the classroom activity. We acknowledge that our instruments cannot reveal whether the attention given to the dictated MWUs also affected the participants’ semantic memory system, which supplies knowledge that is detached from recollections about when and where the items of knowledge were picked up, and which ultimately drives language use and comprehension.

Additional information

Notes on contributors

Frank Boers

Frank Boers is an associate professor at the School of Linguistics and Applied Language Studies of Victoria University of Wellington, New Zealand. His initial research interests were in the field of lexicology (e.g. studies of polysemy) and semantics generally (e.g. studies of metaphor). Most of his more recent research interests, however, were sparked by his long experience as a language teacher and teacher trainer. He now publishes mostly on matters of instructed second language acquisition, especially the teaching of vocabulary and phraseology, and the potential merits of cognitive linguistics in that domain.

Seth Lindstromberg

Seth Lindstromberg works on methodology courses for non-native-speaking, in-service EFL and CLIL teachers. His main study interests are the semantics and usage of English prepositions, the acquisition of multiword units in a foreign language, applications of cognitive linguistics in foreign language teaching, and foreign language teaching methods in general.

June Eyckmans

June Eyckmans is a lecturer of English in the Linguistics Department at Ghent University where she teaches a general applied linguistics course and cognitive linguistic approaches to L2 vocabulary acquisition. She is also affiliated to the Vrije Universiteit Brussels where she teaches a course in advanced linguistics. Her research interests include cognitive approaches to foreign language learning, vocabulary testing, vocabulary acquisition, learner differences, and translation.