Perceptual Consequences of Variability in Native and Non-Native Speech.

BACKGROUND/AIMS: Native speakers often have a difficult time understanding non-native speech, and this challenge is frequently attributed to a more variable signal. While theories and models of general speech perception are grounded in issues of variability, they rarely consider non-native speech. Here, we ask how a specific type of variability (speaking rate) impacts two measures of perception for both native and non-native speech. METHODS: In the present study, one group of listeners transcribed speech, providing a measure of intelligibility. A second group of listeners rated how fluent the speaker was, providing a measure of fluency. RESULTS: The results show that variability in speaking rate correlates with a non-native speaker's intelligibility. However, perceived fluency measures are not predicted by this variability measure. CONCLUSIONS: These results, taken with studies of the range of variability in non-native speech, suggest that variability in non-native speech is not a monolithic construct. Current theories and models of perception can be enhanced by examining non-native speech and how variability in that speech impacts perception.

Speaking rate consistency in native and non-native speakers of English.

Non-native speech differs from native speech in multiple ways. Previous research has described segmental and suprasegmental differences between native and non-native speech in terms of group averages. For example, average speaking rate for non-natives is slower than for natives. However, it is unknown whether non-native speech is also more variable than native speech. This study introduces a method of comparing rate change across utterances, demonstrating that non-native speaking rate is more variable than native speech. These results suggest that future work examining non-native speech perception and production should investigate both mean differences and variability in the signal.

Long-term temporal tracking of speech rate affects spoken-word recognition.

Humans unconsciously track a wide array of distributional characteristics in their sensory environment. Recent research in spoken-language processing has demonstrated that the speech rate surrounding a target region within an utterance influences which words, and how many words, listeners hear later in that utterance. On the basis of hypotheses that listeners track timing information in speech over long timescales, we investigated the possibility that the perception of words is sensitive to speech rate over such a timescale (e.g., an extended conversation). Results demonstrated that listeners tracked variation in the overall pace of speech over an extended duration (analogous to that of a conversation that listeners might have outside the lab) and that this global speech rate influenced which words listeners reported hearing. The effects of speech rate became stronger over time. Our findings are consistent with the hypothesis that neural entrainment by speech occurs on multiple timescales, some lasting more than an hour.

Revisiting Neil Armstrongs Moon-Landing Quote: Implications for Speech Perception, Function Word Reduction, and Acoustic Ambiguity.

Neil Armstrong insisted that his quote upon landing on the moon was misheard, and that he had said one small step for a man, instead of one small step for man. What he said is unclear in part because function words like a can be reduced and spectrally indistinguishable from the preceding context. Therefore, their presence can be ambiguous, and they may disappear perceptually depending on the rate of surrounding speech. Two experiments are presented examining production and perception of reduced tokens of for and for a in spontaneous speech. Experiment 1 investigates the distributions of several acoustic features of for and for a. The results suggest that the distributions of for and for a overlap substantially, both in terms of temporal and spectral characteristics. Experiment 2 examines perception of these same tokens when the context speaking rate differs. The perceptibility of the function word a varies as a function of this context speaking rate. These results demonstrate that substantial ambiguity exists in the original quote from Armstrong, and that this ambiguity may be understood through context speaking rate.

Individual differences in the perception of melodic contours and pitch-accent timing in speech: Support for domain-generality of pitch processing.

Do the same mechanisms underlie processing of music and language? Recent investigations of this question have yielded inconsistent results. Likely factors contributing to discrepant findings are use of small samples and failure to control for individual differences in cognitive ability. We investigated the relationship between music and speech prosody processing, while controlling for cognitive ability. Participants (n = 179) completed a battery of cognitive ability tests, the Montreal Battery of Evaluation of Amusia (MBEA) to assess music perception, and a prosody test of pitch peak timing discrimination (early, as in insight vs. late, incite). Structural equation modeling revealed that only music perception was a significant predictor of prosody test performance. Music perception accounted for 34.5% of variance on prosody test performance; cognitive abilities and music training added only about 8%. These results indicate musical pitch and temporal processing are highly predictive of pitch discrimination in speech processing, even after controlling for other possible predictors of this aspect of language processing.

Distal prosody affects learning of novel words in an artificial language.

The distal prosodic patterning established at the beginning of an utterance has been shown to influence downstream word segmentation and lexical access. In this study, we investigated whether distal prosody also affects word learning in a novel (artificial) language. Listeners were exposed to syllable sequences in which the embedded words were either congruent or incongruent with the distal prosody of a carrier phrase. Local segmentation cues, including the transitional probabilities between syllables, were held constant. During a test phase, listeners rated the items as either words or nonwords. Consistent with the perceptual grouping of syllables being predicted by distal prosody, congruent items were more likely to be judged as words than were incongruent items. The results provide the first evidence that perceptual grouping affects word learning in an unknown language, demonstrating that distal prosodic effects may be independent of lexical or other language-specific knowledge.

Distal rhythm influences whether or not listeners hear a word in continuous speech: support for a perceptual grouping hypothesis.

Due to extensive variability in the phonetic realizations of words, there may be few or no proximal spectro-temporal cues that identify a word's onset or even its presence. Dilley and Pitt (2010) showed that the rate of context speech, distal from a to-be-recognized word, can have a sizeable effect on whether or not a word is perceived. This investigation considered whether there is a distinct role for distal rhythm in the disappearing word effect. Listeners heard sentences that had a grammatical interpretation with or without a critical function word (FW) and transcribed what they heard (e.g., are in Jill got quite mad when she heard there are birds can be removed and Jill got quite mad when she heard their birds is still grammatical). Consistent with a perceptual grouping hypothesis, participants were more likely to report critical FWs when distal rhythm (repeating ternary or binary pitch patterns) matched the rhythm in the FW-containing region than when it did not. Notably, effects of distal rhythm and distal rate were additive. Results demonstrate a novel effect of distal rhythm on the amount of lexical material listeners hear, highlighting the importance of distal timing information and providing new constraints for models of spoken word recognition.

New tests of the distal speech rate effect: examining cross-linguistic generalization.

Recent findings [Dilley and Pitt, 2010. Psych. Science. 21, 1664-1670] have shown that manipulating context speech rate in English can cause entire syllables to disappear or appear perceptually. The current studies tested two rate-based explanations of this phenomenon while attempting to replicate and extend these findings to another language, Russian. In Experiment 1, native Russian speakers listened to Russian sentences which had been subjected to rate manipulations and performed a lexical report task. Experiment 2 investigated speech rate effects in cross-language speech perception; non-native speakers of Russian of both high and low proficiency were tested on the same Russian sentences as in Experiment 1. They decided between two lexical interpretations of a critical portion of the sentence, where one choice contained more phonological material than the other (e.g., /str'na/ "side" vs. /str'na/ "country"). In both experiments, with native and non-native speakers of Russian, context speech rate and the relative duration of the critical sentence portion were found to influence the amount of phonological material perceived. The results support the generalized rate normalization hypothesis, according to which the content perceived in a spectrally ambiguous stretch of speech depends on the duration of that content relative to the surrounding speech, while showing that the findings of Dilley and Pitt (2010) extend to a variety of morphosyntactic contexts and a new language, Russian. Findings indicate that relative timing cues across an utterance can be critical to accurate lexical perception by both native and non-native speakers.