The role of perceptual salience in the L2 acquisition sequence of the Arabic construct state


University of Mi, Ann Arbor


Sun, 09/25/2016 - 12:00am


CMES Faculty Travel Award

Mahmoud Azaz, Assistant Professor MENAS and SLAT

Arabic SLA Conference


The effects of perceptual salience, the ease in which grammatical features can be noticed in the input, continues to be a debated topic in acquisition sequence and order studies (Goldschneider & DeKeyser, 2001; DeKeyser 2003; N. Ellis & Larsen-Freeman, 2006). Specifically, this study asks to what extent the high salience of head direction and the low salience of definiteness determines the acquisition sequence of the definite Arabic construct state (CS), an under researched phenomenon in Arabic SLA. Mapping English definite possessive s-genitive constructions (the boy’s shirt) and compounds (the love letter) to their equivalent Arabic CSs (qamiiṣ-u al-walad-i and risaalat-u al ḥubb-i) entails reordering the head (al-muḍaaf) and the complement (al-muḍaaf ilihi) and the position of the definite marker. In line with Goldschneider and DeKeyser (2001), head-complement reordering (i.e., head direction resetting) is more salient than definiteness adjustment because the head and the complement have more phones and syllables than the monosyllabic definite marker (al-) that never occurs independently. Further, structural dissimilarity in head direction in the target constructions contributes to its salience.

Three English-speaking learner groups (beginners, n=15; intermediate, n=15; and advanced, n=15) participated in this study. The beginners were tested after about 28 hours of instruction, during which they did not receive explicit instruction of the target construction. Data were collected using a sentence completion task, prompted with pictures, which elicited definite CSs while the participants’ focus was on meaning. A total of 18 target items, scrambled with 18 distractors, were equally distributed into CSs that were equivalent to definite possessive s-genitives and those equivalent to compounds.

Results showed a strong effect for salience in the early stage of acquisition; the beginners were to a great extent(64.70%) sensitive to contrasting head direction, although they did not receive explicit instruction on the Arabic CS. However, errors in definiteness brought their overall accuracy down to 29.41%. The intermediate group patterned the beginning group. The advanced group performed at ceiling in terms of head direction (100%), but continued to have issues with definiteness that brought their accuracy score down to 86.86%. To further scrutinize the sequence of acquisition, a multi-layered analysis was conducted. A across-group comparison was made between the accuracy scores that entailed head direction resetting and definiteness adjustment and the means of the left-headed responses that entailed only head direction resetting. No differences were found between the mean scores of the beginning group’s left-headed responses and the intermediate group’s overall accuracy in a one-way ANOVA: F(1,29) = .506, p = .483. Contrastively, there was a clear difference between the intermediate group’s left-headed responses and the advanced group’s overall accuracy: F(1,29) = 106.22, p < .001.

The overall results suggest that in the absence of explicit instruction early L2 learners of Arabic were sensitive the salient feature of head direction. Further, the acquisition of the definite CS by English-speaking learners proceeds in two gradual phases: Early on, head direction is acquired. Only later on definiteness is adjusted. Implications in terms of incidental/implicit syntactic learning and the effect of salience in the acquisition sequence are further discussed.