This study focuses on two experiments conducted with Eastern and Western Balochi speakers. In Eastern Balochi, voiceless stops have aspirated features, but in Western Balochi, they are unaspirated. Eighty-four native speakers of both dialects of Balochi participated in this study. Participants of the first experiment produced words of their L1 in a picture naming task, and those of the second experiment read words in English. VOTs of the L1 and L2 voiced stops elicited from recordings of the productions. Results show that speakers of Western Balochi transfer their L1 negative VOTs to L2 English-voiced stops. However, Eastern Balochi speakers produce English-voiced stops with VOTs, significantly different from their L1 VOTs. Though they could not produce English-voiced stops with native-like accuracy, they produced English stops with significantly longer pre-voicing duration than their L1-voiced stops. Therefore, the study concludes that speakers of those languages with stops with negative VOT ranges face more difficulty acquiring L2 voiced stops of short-lag positive VOTs than those learners whose L1 does not have such stops. The speech learning model is used in this study to analyze all results.

An important issue in the recent L2 literature is whether adult learners can perceive a gradient phonetic difference between L1 and the corresponding L2 phonemes. This leads us to further judge whether such learners can acquire allophonic contrast in L2. The same issue has been addressed in the current study with reference to Kuwaiti adult learners of English. The aim of this study is to analyze velar stops /k, g/ as perceived and produced by Kuwaiti Arabic (KA) learners of English. In this study, comprehension is measured by a perception test, and pronunciation is measured by obtaining readings of voice onset time (VOT) in Praat software. The velar stops will also be analyzed using the framework of the classic version of optimality theory (OT). OT became one of the major generative frameworks in the field of generative phonology. OT is a constraint-based theory of phonology advanced by Prince and Smolensky. In the perception test, participants were asked to listen to words containing target sounds, which were "keys," "skis," and "geese," and note on paper which English word they had heard. In the discrimination task, these words were presented in pairs, and the participants were asked to say whether they had heard a single word twice or two words together. In the production test, they read from a list of words, including "keys," "skis," and "geese," and their productions were recorded. A large group of Kuwaiti English learners perceived and produced these velar stops of English. Their discrimination of voiced and voiceless velar stops was excellent, but their identification of [g] was weak. Their identification of the voiceless velar stop /k/ was also native-like. In production, they were native-like in aspirated [k h ]. In the production of the unaspirated voiceless velar stop [k], they were not native-like, but they had developed an understanding of this allophone of English. In [g], some students were native-like, some were still learning, and some had only relied on L1 transfer.