Fine-grained Phonetic Details Research Articles

Spoken word recognition in a second language: The importance of phonetic details.

Spoken word recognition depends on variations in fine-grained phonetics as listeners decode speech. However, many models of second language (L2) speech perception focus on units such as isolated syllables, and not on words. In two eye-tracking experiments, we investigated how fine-grained phonetic details (i.e. duration of nasalization on contrastive and coarticulatory nasalized vowels in Canadian French) influenced spoken word recognition in an L2, as compared to a group of native (L1) listeners. Results from L2 listeners (English-native speakers) indicated that fine-grained phonetics impacted the recognition of words, i.e. they were able to use nasalization duration variability in a way similar to L1-French listeners, providing evidence that lexical representations can be highly specified in an L2. Specifically, L2 listeners were able to distinguish minimal word pairs (differentiated by the presence of phonological vowel nasalization in French) and were able to use variability in a way approximating L1-French listeners. Furthermore, the robustness of the French "nasal vowel" category in L2 listeners depended on age of exposure. Early bilinguals displayed greater sensitivity to some ambiguity in the stimuli than late bilinguals, suggesting that early bilinguals had greater sensitivity to small variations in the signal and thus better knowledge of the phonetic cue associated with phonological vowel nasalization in French, similarly to L1 listeners.

An analysis of listening errors by Korean EFL learners from self-paced passage dictation

In this study, listening errors by Korean EFL learners are comprehensively analyzed from self-paced passage dictation tasks. Fifty-five Korean EFL learners participated in the study. Listeners were asked to write down dictation passages as accurately as possible, while listening to the audio as much as they needed. The results show that (i) low-proficiency learners tend to misperceive longer phrases than high-proficiency learners, (ii) function words are more often omitted or misheard than content words, and (iii) low-proficiency learners have more difficulties with content words than high-proficiency learners do. Most frequent suffix errors were omissions of past or plural suffixes. Among the function words, the most frequent errors were found with auxiliary contractions, infinitive marker to, and articles, mostly in the environment of linking and elision. It is also shown that C-V linking, C-C linking, and elision are the primary sources for the most frequent errors. C-V linking led to errors in correctly locating the word boundary, while C-C linking and elision resulted in omission. These errors show that Korean EFL listeners have difficulties in detecting fine-grained phonetic details to the extent that native speakers can do.

CONCEPTUALLY CUED PERCEPTUAL CATEGORIZATION IN ADULT L2 LEARNERS

AbstractPrevious studies attest that early bilinguals can modify their perceptual identification according to the fine-grained phonetic detail of the language they believe they are hearing. Following Gonzales et al. (2019), we replicate the double phonemic boundary effect in late learners (LBs) using conceptual-based cueing. We administered a forced choice identification task to 169 native English adult learners of Spanish in two sessions. In both sessions, participants identified the same /b/-/p/ voicing continuum, but language context was cued conceptually using the instructions. The data were analyzed using Bayesian multilevel regression. Learners categorized the continuum in a similar manner when they believed they were hearing English. However, when they believed they were hearing Spanish, “voiceless” responses increased as a function of L2 proficiency. This research demonstrates the double phonemic boundary effect can be conceptually cued in LBs and supports accounts positing selective activation of independent perception grammars in L2 learning.

Centering diphthongization in Hui Chinese dialects

This paper reports centering diphthongization in three Hui Chinese dialects. It argues that centering diphthongization serves as clear evidence that monophthongs and diphthongs are not a dichotomy, but a continuum, and thus supports a dynamic theory of vowel production (Hu, 2017). Centering diphthongs in Hui dialects occur as an intermediate vowel category between monophthongs and diphthongs. Fine-grained phonetic details reveal that the process of centering diphthongization is gradient in the three Hui dialects. The Yi county dialect is at an early stage: centering diphthongization is essentially a phonetic process of adding an additional neutralized [ɐ]-like production to the distinctive vowel element, namely [i u y e o] > [i:ɐ u:ɐ y:ɐ e:ɐ o:ɐ], respectively. The Qimen dialect is at a phonologically developed stage: monophthongs and diphthongized vowels contrast in dynamics, namely the static [i u y] versus the dynamic [i:ɐ u:ɐ y:ɐ], but diphthongized vowels have a temporal structure different to that of plain diphthongs. The Xiuning dialect is at a final stage: monophthongs and diphthongized vowels contrast in dynamics, and diphthongized vowels exhibit a temporal organization similar to that of plain diphthongs.This paper reports centering diphthongization in three Hui Chinese dialects. It argues that centering diphthongization serves as clear evidence that monophthongs and diphthongs are not a dichotomy, but a continuum, and thus supports a dynamic theory of vowel production (Hu, 2017). Centering diphthongs in Hui dialects occur as an intermediate vowel category between monophthongs and diphthongs. Fine-grained phonetic details reveal that the process of centering diphthongization is gradient in the three Hui dialects. The Yi county dialect is at an early stage: centering diphthongization is essentially a phonetic process of adding an additional neutralized [ɐ]-like production to the distinctive vowel element, namely [i u y e o] > [i:ɐ u:ɐ y:ɐ e:ɐ o:ɐ], respectively. The Qimen dialect is at a phonologically developed stage: monophthongs and diphthongized vowels contrast in dynamics, namely the static [i u y] versus the dynamic [i:ɐ u:ɐ y:ɐ], but diphthongized vowels have a temporal structure different to that o...

Fricative vowels as an intermediate stage of vowel apicalization

Abstract Diphthongization and apicalization are two commonly detected phonetic and/or phonological processes for the development of high vowels, with the process of apicalization being of particular importance to the phonology of Chinese dialects. This paper describes acoustics and articulation of fricative vowels in the Suzhou dialect of Wu Chinese. Acquiring frication initiates the sound change. The production of fricative vowels in Suzhou is characterized by visible turbulent frication from the spectrograms, and a significant lower Harmonics-to-Noise Ratio vis-à-vis the plain counterparts. The acoustic study suggests that spectral characteristics of fricative vowels play a more important role in defining the vowel contrasts. The fricative high front vowels have comparatively greater F1 and smaller F2 and F3 values than their plain counterparts, and in the acoustic F1/F2 plane, the fricative vowels are located in an intermediate position between their plain and apical counterparts. The articulatory study revealed that that not only tongue dorsum but also tongue blade are involved in the production of fricative high front vowels in Suzhou. Phonologically, plain high front vowels, fricative high front vowels, and apical vowels distinguish in active place of articulation, namely being anterodorsal, laminal, and apical respectively; and frication becomes a concomitant and redundant feature in the production of fricative or apical vowels. It is concluded that the fine-grained phonetic details suggest that the fricative high front vowels in Suzhou is at an intermediate stage of vowel apicalization in terms of both acoustics and articulation.

Non-native perception of regionally accented speech of competing talkers

Noisy conditions are challenging to non-native listeners who typically underperform relative to native listeners when attending to several competing talkers. We examine whether non-native listeners can utilize dialect-related cues in the target and in the masking speech, even if they do not reach the proficiency level of the native listeners. Our previous work with highly proficient Indonesian-English bilinguals (Fox et al., 2014) found that their performance differed markedly from native English listeners when the speech levels of the competing talkers were equal (0 dB SNR). We hypothesized that bilinguals cannot effectively separate utterances at 0 dB SNR due to a lack of a sufficient contrast in the voice levels of competing talkers, which may reduce their ability to benefit from the phonetic-acoustic details in order to “follow” a particular talker. The current study sought to replicate and validate earlier findings with a different group of bilinguals. The same experiment was conducted with Korean-English bilinguals. The results were replicated, providing further evidence that the ability to benefit from fine-grained phonetic details in regional accents declines for non-native listeners when the speech levels of the competing talkers are equal. Discussion will focus on defining the nature of the non-native speech processing deficit.

Musical training and the perception of phonetic detail in a shadowing task

Phonetic convergence, or shadowing, is the phenomenon in which people unintentionally and temporarily change phonetic details of their speech to sound more similar to another talker. Previous research has examined effects of the shadower and of the target speaker. In the current study, we examine listener contributions on the perception of phonetic convergence. Two hundred and sixty listeners completed an AXB perception task in which they were asked to determine whether the first or third stimulus was a better imitation of the middle stimulus. Listeners provided information about previous instrumental and vocal training. Preliminary examination of the results reveals that listeners with more musical training (in the form of instrumental or vocal experience) lead to better accuracy. Results from the study suggest that there is a link between musical training and detection of fine-grained phonetic details in speech.

Vowel Nasalisation in Scottish Gaelic: The Search for Paradigm Uniformity Effects in Fine-Grained Phonetic Detail

According to the modular feedforward architecture of grammar, the phonetics is sensitive only to the output of the phonology and is thus blind to morphological or lexical conditioning (Pierrehumbert 2002). However, this prediction is challenged by claims that fine-grained phonetic detail may display paradigm uniformity (PU) effects (Steriade 2000). In the present study I search for phonetic PU effects in vowel nasalisation in Scottish Gaelic by investigating alternating items in which a nasalising environment is removed by a morpho(phono)logical process known as lenition, which replaces initial [m] with [v] under certain morphosyntactic conditions.In vowels following initial [m], a clear distinction is found between (i) categorical phonological nasalisation, which may be subject to lexically conditioned blocking and which displays overapplication in lenited forms, and (ii) gradient phonetic nasalisation, which applies in those items where categorical phonological nasalisation fails to occur and which disappears completely in lenited forms. The differing patterns displayed by these two types of nasalisation fit neatly with the predictions of a modular architecture, in which categorical phonology has direct access to morphological information but gradient phonetics does not, and I conclude that non-modular architectures such as Exemplar Theory are not the correct explanation for putative PU effects.

Generalization of Phonetic Detail: Cross-Segmental, Within-Category Priming of VOT.

The current study examined whether fine-grained phonetic detail (voice onset time (VOT)) of one segment (/p/ or /k/) generalizes to a different segment (/t/) within the same natural class. Two primes were constructed to exploit the natural variation of VOT: a velar stop followed by a high vowel (keen) resulting in a naturally long VOT and a labial stop followed by a low vowel (pan) resulting in a naturally shorter VOT. Two experiments were conducted, one in which the speakers produced both the prime and the target, and a second in which the speakers heard the primes and then produced the targets. In Experiment 1, VOTs for initial /t/ were shorter following pan than following keen. In Experiment 2 where participants heard the primes, priming was found only when the primes had unexpected relative VOT values (short for keen and long for pan). These results provide evidence for cross-segmental generalization of phonetic detail and also suggest that natural, within-category variability is encoded during language processing.

Loanword adaptation: Phonological and cognitive issues

AbstractIn this paper, I focus on loanwords, and more specifically on the process known as loanword adaptation, namely the process by which a word’s original phonetics is altered to phonologically fit the borrowing language. More precisely, I seek to determine whether release bursts after the word-final coda in an English input might be responsible for final vowel epenthesis in the corresponding Korean output. To test this hypothesis, I compiled an original database listing English words that entered Korean until very recently. The analysis of that database actually supports the hypothesis. In a final conclusion, I discuss what this demonstration tells us about loanword adaptation and the process of borrowing from a psycholinguistic point of view. In particular, I emphasize the importance of both perception and native phonology in the process of loanword adaptation and conclude that Korean speakers seem indeed to be subject to the ‘perceptual magnet effect’, seeking (phonological) prototypes from various acoustic cues, and that so-called fine-grained phonetic detail has to be explored for a better understanding of phonological processes from a cognitive point of view.

Vowel development in children with Down and Williams syndromes

Down syndrome (DS) and Williams syndrome (WS) are genetic disorders resulting from different types of genetic errors. While both disorders lead to phonological and speech motor deficits, particularly little is known about vowel production in DS and WS. Recent work suggests that impaired vowel articulation in DS likely contributes to the poor intelligibility of DS speech. Developmental delays in temporal vowel structure and pitch control have been found in children with WS when compared to their chronological matches. Here, we analyze spontaneous speech samples produced by British children with DS and WI and compare them with typically developing children from the same geographic area in Southern England. We focus on the acquisition of fine-grained phonetic details, asking if children with DS and WS are able to synchronize the phonetic and indexical domains while coping with articulatory challenges related to their respective syndromes. Phonetic details pertaining to the spectral (vowel-inherent spectral change) and indexical (regional dialect) vowel features are examined and vowel spaces are derived from formant values sampled at multiple temporal locations. Variations in density patterns across the vowel space are also considered to define the nature of the acoustic overlap in vowels related to each syndrome.

Effects of acoustic–phonetic detail on cross-language speech production

Abstract Nonnative sounds and sequences are systematically adapted in both perception and production. For example, American English speakers often modify illegal word-initial clusters by inserting a vocalic transition between the two consonants (e.g., (/bdagu/ → [bədagu]). Previous work on such modifications has for the most part focused on relatively abstract properties of the nonnative structures, such as their phonemic content and whether they conform to sonority sequencing principles. The current study finds that fine-grained phonetic details of the stimulus can be equally important for predicting cross-language production patterns. Several acoustic–phonetic properties were manipulated to create stimulus variants that are phonemically identical (i.e., exhibit non-contrastive variation) in the target language (Russian). In a shadowing experiment, English speakers’ correct productions and detailed error patterns were significantly modulated by the acoustic manipulations. The results highlight the role of perception in accounting for cross-language production, and establish limits on the perceptual repair of nonnative sound sequences by phonetic decoding.

Perceptual Recovery from Consonant-Cluster Simplification in Korean Using Language-Specific Phonological Knowledge

Two experiments examined whether perceptual recovery from Korean consonant-cluster simplification is based on language-specific phonological knowledge. In tri-consonantal C1C2C3 sequences such as /lkt/ and /lpt/ in Seoul Korean, either C1 or C2 can be completely deleted. Seoul Koreans monitored for C2 targets (/p/ or / k/, deleted or preserved) in the second word of a two-word phrase with an underlying /l/-C2-/t/ sequence. In Experiment 1 the target-bearing words had contextual lexical-semantic support. Listeners recovered deleted targets as fast and as accurately as preserved targets with both Word and Intonational Phrase (IP) boundaries between the two words. In Experiment 2, contexts were low-pass filtered. Listeners were still able to recover deleted targets as well as preserved targets in IP-boundary contexts, but better with physically-present targets than with deleted targets in Word-boundary contexts. This suggests that the benefit of having target acoustic-phonetic information emerges only when higher-order (contextual and phrase-boundary) information is not available. The strikingly efficient recovery of deleted phonemes with neither acoustic-phonetic cues nor contextual support demonstrates that language-specific phonological knowledge, rather than language-universal perceptual processes which rely on fine-grained phonetic details, is employed when the listener perceives the results of a continuous-speech process in which reduction is phonetically complete.

Aerodynamic modeling for concatenative speech synthesis.

Listeners can perceive and use a wide array of fine-grained phonetic details, including the detailed coarticulatory influences of adjacent sounds, when perceiving speech. Details like anticipatory nasalization in can, for example, potentially provide the listener with a rich network of informative cues and are a key to understanding listeners’ ability to disambiguate speech sounds from seemingly ambiguous input. Unfortunately, these coarticulatory cues are generally missing or contradictory in the output of speech synthesis systems. These systems work by concatenating variable-length sound units chosen from a large database of recorded speech. Units are chosen to minimize two functions: the cost of aligning a particular unit with the desired speech output (target cost) and the cost of adjoining the next sound to the most recently selected unit (join cost). Generally, these costs are calculated using features which can be automatically extracted from the acoustic speech signal. A unit selection database is created, automatically segmented and automatically labeled with nasal and oral airflow feature vectors. These aerodynamic features are used as a proxy for articulatory information in the calculation of join and cost functions. Listeners’ mean opinion scores are obtained on output from this system and a baseline acoustic system for comparison.

Phonetic interpretation: Papers in laboratory phonology 6 (review)

Reviewed by: Phonetic interpretation: Papers in laboratory phonology 6 ed. by John Local, Richard Ogden, and Rosalind Temple Maria-Josep Solé Phonetic interpretation: Papers in laboratory phonology 6. Ed. by John Local, Richard Ogden, and Rosalind Temple. Cambridge: Cambridge University Press, 2004. Pp. xiv, 402. ISBN 0521824028. $85 (Hb). This new volume of the ‘Papers in laboratory phonology’ series features a collection of papers presented at the sixth Laboratory Phonology meeting in York in 1998, and published in 2004, five years later. This volume focuses on phonetic interpretation. The term phonetic interpretation has traditionally assumed the conversion of abstract, discrete, and invariant phonological representations to continuous and varying speech signals. A number of papers in the volume furnish evidence that forces us to reassess the nature of phonological representations, as John Coleman rightly points out. The papers show that fine phonetic detail—which traditionally has played no role in phonological distinctions—is used by listeners to identify words (Sarah Hawkins [End Page 438] and Noël Nguyen), is used differently in dialects to indicate segmental contrasts (Paul Carter), and is exploited by speakers to identify themselves with a particular social group (Gerald J. Docherty). Consequently, fine-grained phonetic detail must be part of the lexical representation and under the control of the speaker. Other papers in the volume provide evidence on the role of frequency of occurrence (token frequency) on the pronunciation of segments (Richard Wright), and the role of frequency of the pattern in the lexicon (type frequency) on speech perception and on well-formedness judgments (Jennifer Hay, Janet Pierrehumbert, and Mary E. Beckman). These findings on the role of phonetic detail and frequency in speech processing and speech production cannot be accounted for by abstract phonological representations stripped of phonetic detail and suggest probabilistic models of representation of phonological categories such as those put forward by exemplar models or usage-based models (e.g. Johnson 1997, Bybee 2001, Pierrehumbert 2002). If this is so and phonetic variability itself is part of the make-up of the category, the focus of the book should not be on the phonetic interpretation of phonological representations, but rather on how speakers form phonological categories and how phonological categories structure phonetic variation. The sixteen papers are organized in four thematic sections each followed by a rigorous and comprehensive commentary by a renowned expert. The ‘Introduction’ written by the editors is an excellent review of the aims, practices, and research agenda of laboratory phonology, providing a common conceptual framework for the diverse papers in the volume. Beckman and Pierrehumbert’s paper provides a background to the first section, ‘Phonological representation and the lexicon’. They review key issues in category formation, the arbitrariness of the linguistic sign, duality of patterning, the organization of the lexicon, and lexical retrieval. They couch these issues in terms of connectionist models (Dell 1988) and usage-based models (Bybee 2000). It is surprising though that they revisit long-standing notions without acknowledging the original authors. Hawkins and Nguyen present perceptual evidence that phonetic properties of onset /1/ (consonant duration, F0 but not F2 frequency) can be used by speakers to predict the voicing of the coda obstruent in the same syllable (e.g. led vs. let). The two main findings of their study, that phonetic exponents of segmental contrasts can be present in nonadjacent segments and that the listeners show sensitivity to fine-grained acoustic information, lead them to suggest a word-based model of lexical access and that lexical representations must include fine-phonetic detail. Wright’s paper demonstrates that the pronunciation of individual lexical items is affected by the number and lexical frequency of the word’s lexical neighbors. Thus, words from high-density neighborhoods and with low relative frequency use a more expanded vowel space. The expansion of the vowel space is such that only peripheral vowels—which can move to more extreme positions without diminishing the vowel contrasts—become more dispersed whereas the others remain fairly unchanged. This result is an interesting addition to the literature on the phonetic dimensions that increase (or decrease) perceptual contrast among words...

The modulation of lexical competition by segment duration

In an eye-tracking study, we examined how fine-grained phonetic detail, such as segment duration, influences the lexical competition process during spoken word recognition. Dutch listeners' eye movements to pictures of four objects were monitored as they heard sentences in which a stop-initial target word (e.g., pijp "pipe") was preceded by an [s]. The participants made more fixations to pictures of cluster-initial words (e.g., spijker "nail") when they heard a long [s] (mean duration, 103 msec) than when they heard a short [s] (mean duration, 73 msec). Conversely, the participants made more fixations to pictures of the stop-initial words when they heard a short [s] than when they heard a long [s]. Lexical competition between stop- and cluster-initial words, therefore, is modulated by segment duration differences of only 30 msec.

Prosodic strengthening and featural enhancement: Evidence from acoustic and articulatory realizations of /ɑ,i/ in English

In this study the effects of accent and prosodic boundaries on the production of English vowels (/a,i/), by concurrently examining acoustic vowel formants and articulatory maxima of the tongue, jaw, and lips obtained with EMA (Electromagnetic Articulography) are investigated. The results demonstrate that prosodic strengthening (due to accent and/or prosodic boundaries) has differential effects depending on the source of prominence (in accented syllables versus at edges of prosodic domains; domain initially versus domain finally). The results are interpreted in terms of how the prosodic strengthening is related to phonetic realization of vowel features. For example, when accented, /i/ was fronter in both acoustic and articulatory vowel spaces (enhancing [-back]), accompanied by an increase in both lip and jaw openings (enhancing sonority). By contrast, at edges of prosodic domains (especially domain-finally), /i/ was not necessarily fronter, but higher (enhancing [+high]), accompanied by an increase only in the lip (not jaw) opening. This suggests that the two aspects of prosodic structure (accent versus boundary) are differentiated by distinct phonetic patterns. Further, it implies that prosodic strengthening, though manifested in fine-grained phonetic details, is not simply a low-level phonetic event but a complex linguistic phenomenon, closely linked to the enhancement of phonological features and positional strength that may license phonological contrasts.