Clear speech in Tashlhiyt Berber: The perception of typologically uncommon word-initial contrasts by native and naive listeners

Social factors condition consonant separation in Moroccan Arabic word-initial clusters: Amazigh-Arabic bilingualism, region, and speaking style

Moroccan Arabic is a regional variety of spoken Arabic that contains linguistic features, which make it distinct from other dialects. Moroccan Arabic, for instance, contains a larger number of words with consonant clusters, a result of a historical vowel reduction process, compared to other Eastern varieties. Even within Moroccan Arabic, there is variation across speakers that is relatively understudied. The present study investigates across- and within-speaker factors that condition variation in the duration of acoustic separation between adjacent consonants in word-initial clusters in Moroccan Arabic. Three factors are examined: speaking style (clear vs fast speech), speaker region of origin (the Easternmost region of Morocco vs other regions), and speaker first language (Moroccan Arabic vs an Amazigh language). It is determined that (1) speakers produce more frequent, and longer when present, consonant separation in clear speech; (2) Amazigh-Arabic bilingual speakers produce higher rates and longer durations of consonant separations between adjacent consonants; and (3) regional dialect variation is also observed: speakers of Easternmost Moroccan Arabic produce more frequent and longer consonant separations than speakers from other Moroccan regions. Results are discussed for implications of speech production models as well as for theories of sound change.

Learning a language with vowelless words

Vowelless words are exceptionally typologically rare, though they are found in some languages, such as Tashlhiyt (e.g., fkt 'give it'). The current study tests whether lexicons containing tri-segmental (CCC) vowelless words are more difficult to acquire than lexicons not containing vowelless words by adult English speakers from brief auditory exposure. The role of acoustic-phonetic form on learning these typologically rare word forms is also explored: In Experiment 1, participants were trained on words produced in either only Clear speech or Casual speech productions of words; Experiment 2 trained participants on lexical items produced in both speech styles. Listeners were able to learn both vowelless and voweled lexicons equally well when speaking style was consistent for participants, but learning was lower for vowelless lexicons when training consisted of variable acoustic-phonetic forms. In both experiments, responses to a post-training wordlikeness ratings task containing novel items revealed that exposure to a vowelless lexicon leads participants to accept new vowelless words as acceptable lexical forms. These results demonstrate that one of the typologically rarest types of lexical forms - words without vowels - can be rapidly acquired by naive adult listeners. Yet, acoustic-phonetic variation modulates learning.

Comparing perception of L1 and L2 English by human listeners and machines: Effect of interlocutor adaptationsa)

Speakers tailor their speech to different types of interlocutors. For example, speech directed to voice technology has different acoustic-phonetic characteristics than speech directed to a human. The present study investigates the perceptual consequences of human- and device-directed registers in English. We compare two groups of speakers: participants whose first language is English (L1) and bilingual L1 Mandarin-L2 English talkers. Participants produced short sentences in several conditions: an initial production and a repeat production after a human or device guise indicated either understanding or misunderstanding. In experiment 1, a separate group of L1 English listeners heard these sentences and transcribed the target words. In experiment 2, the same productions were transcribed by an automatic speech recognition (ASR) system. Results show that transcription accuracy was highest for L1 talkers for both human and ASR transcribers. Furthermore, there were no overall differences in transcription accuracy between human- and device-directed speech. Finally, while human listeners showed an intelligibility benefit for coda repair productions, the ASR transcriber did not benefit from these enhancements. Findings are discussed in terms of models of register adaptation, phonetic variation, and human-computer interaction.

Linguistic disparities in cross-language automatic speech recognition transfer from Arabic to Tashlhiyt

Tashlhiyt is a low-resource language with respect to acoustic databases, language corpora, and speech technology tools, such as Automatic Speech Recognition (ASR) systems. This study investigates whether a method of cross-language re-use of ASR is viable for Tashlhiyt from an existing commercially-available system built for Arabic. The source and target language in this case have similar phonological inventories, but Tashlhiyt permits typologically rare phonological patterns, including vowelless words, while Arabic does not. We find systematic disparities in ASR transfer performance (measured as word error rate (WER) and Levenshtein distance) for Tashlhiyt across word forms and speaking style variation. Overall, performance was worse for casual speaking modes across the board. In clear speech, performance was lower for vowelless than for voweled words. These results highlight systematic speaking mode- and phonotactic-disparities in cross-language ASR transfer. They also indicate that linguistically-informed approaches to ASR re-use can provide more effective ways to adapt existing speech technology tools for low resource languages, especially when they contain typologically rare structures. The study also speaks to issues of linguistic disparities in ASR and speech technology more broadly. It can also contribute to understanding the extent to which machines are similar to, or different from, humans in mapping the acoustic signal to discrete linguistic representations.