Splitting Unsplittable Foreign Words in Casual Speech by EFL Arab Learners

Abstract

This study sheds light on a novel mispronunciation problem in L2 where Arabic-speaking EFL learners split unsplittable foreign long words in the flow of speech into two sub-parts and the factors involved in this faulty word segmentation. A sample of 15 unsplittable foreign long words segmented by 74 Arab college students was analyzed. Results showed that Arab learners split words to two parts as in Skype > Sky + pe, Kaspersky > Kasper + sky, Swarovski > Swaro+viski, Google > Go + gil, vegetable > vege + table, marshmallow > Marsh + mello, Michigan > Mit + shigan, Wednesday > Wednes + day, manipulated > manu + plated and so on. In segmenting long words, Arab learners often rely on the words’ written form, treating unfamiliar long words as consisting of familiar parts and pronouncing them as if they were two words, with a slight pause between both parts, especially in the case of segments that resemble known English words as sky, table, go, day, marsh. They rearrange consonant clusters based on their Arabic (L1) phonotactic constraints, insert a vowel to break the clusters, and stress the penultimate syllable in the second part. Some faulty word segmentation in the sample is based on cross-linguistic lexical associations, where segments evoke meaningful words in their native language (Arabic) as Swaro سوار meaning (bracelet). Learners intuitively reconstruct unfamiliar words using analogies from both English and Arabic. Their pattern of phonological segmentation is driven by both orthographic influence and phonotactic constraints in Arabic. Learners try to make sense of unfamiliar phonological forms using Arabic phonology, English orthography, and semantic associations. To enable Arab learners to pronounce long words without segmenting them, this study recommends awareness-raising of orthographic mismatches, phonetic awareness training, stress and rhythm drills, listening to native speakers and shadowing their pronunciation, pause, juncture and boundaries, metacognitive strategy training, contrastive analysis and cross-linguistic awareness, and interactive activities. A variety of technologies can be used in pronunciation practice as well. Results of the data analysis, sources of faulty word segmentation and recommendation for teaching and practice are described in detail.