An Empirical Investigation into the Learning of Novel Collocations Using Alliteration as a Mnemonic Device

The central question addressed in this experiment is to what extent phonological patterns, in particular alliteration, aid the recall and retention of novel word-pairs for learners of English. The research builds on previous findings which show a mnemonic advantage for collocations that have phonological patterns, compared to equivalent word strings with no phonological overlap. This advantage appears in both free- and cued-recall tests, and across a variety of temporal intervals (up to two weeks). Much of the prior research has drawn participant samples from a Dutch L1 speaking population. Furthermore, these studies have mainly used target items deemed to be familiar to the participants, meaning that only the form (or spelling) is tested.

This experiment is motivated by the need to question if these previous findings generalise to a population whose L1 phonological constructs are different from those of Dutch L1 speakers. The purpose is to test if Japanese L1 speakers have a different perception of alliteration, and if so, whether this impacts on their learning behaviour. A further aim is to investigate whether the mnemonic effect applies to the meaning, not only the form, of novel target items. The treatment phase incorporates a dictation activity followed by a study phase, using unfamiliar Adjective + Noun collocations as the target items. The testing instruments that were used to measure retention include an immediate free-recall test of the written forms, followed by a three-week delayed-recall test of both form and meaning. Overall, the findings indicate that alliteration does confer a mnemonic advantage, albeit with a modest effect size, but the cognitive challenge of learning new material appears to mitigate any robust mnemonic effects. In answer to the central research question, it can be inferred that alliteration is a useful pedagogical tool for helping language learners recall and retain multi-word strings.