*A practice conference talk

Abstract
This study investigates L2 Spanish rhotic production in learners enrolled in first-semester and fourth-semester courses, specifically addressing the effects that the different ways to produce American English rhotics (retroflex and bunched) have on the production of Spanish taps and trills. It also addresses the influence that the phonological environment producing taps in English has on the acquisition of Spanish taps. The research questions that drove this study are the following:

1. Does Manner of American English rhotic articulation (i.e., retroflex or bunched) affect the facilitation of Spanish rhotic production?

2. Does the phonological environment that governs taps in English affect the accuracy in Spanish rhotic production?

3. Are these hypothesized effects evident in more advanced learners as well as beginners?

Forty-eight students enrolled in first-semester Spanish foreign language classes and thirty-five students enrolled in fourth-semester Spanish foreign language classes were recorded while reading a Spanish text (five short paragraphs) aloud containing thirty-two intervocalic taps and four intervocalic trills. Of the taps, nineteen occurred in environments that would produce taps in English and thirteen taps occurred in other environments. The higher proficiency group also read a second text that contained forty-two intervocalic trills.

Results from multiple linear regressions show that English rhotic articulation alone is a significant predictor of trill accuracy. English rhotic articulation is also a predictor of tap accuracy when controlling for amount of Spanish exposure for beginning learners. These effects, however, disappear altogether in the more advanced learners. Concerning
the effect of the phonological environment producing taps in English on the production of Spanish rhotics, results from paired samples t-tests show that both the first-semester and the fourth-semester groups produced accurate taps significantly more in Spanish words that have the same phonological environment that produces taps in English.

These results suggest that learners who employ bunched-like articulations are at a slight disadvantage to their retroflex-like producing counterparts at the point when they notice a contrast between English and Spanish rhotics. This disadvantage subsequently disappears as exposure to Spanish increases and is non-existent in the more advanced learners. The phonological environment that produces taps in English influences the accuracy of Spanish taps for both proficiency levels and is therefore a stronger influence on Spanish rhotic accuracy. This implies that less conscious and more abstract phonological phenomena have a stronger and longer lasting impact on Spanish rhotic production than phonetic factors.