Perception-production Link Research Articles

Perception-production link mediated by position in the imitation of Korean nasal stops.

This study explores how perceptual cues in two positions influence imitation of Korean nasal stops. As a result of initial denasalization, nasality cues are secondary in the initial position but primary in the medial position. Categorization and imitation tasks using CV (consonant-vowel) and VCV (vowel-consonant-vowel) items on a continuum from voiced oral to nasal stops were completed by 32 Korean speakers. Results revealed categorical imitation of nasality medially, whereas imitation was gradient or minimal initially. Furthermore, individuals requiring stronger nasality cues to categorize a nasal sound produced greater nasality in imitation. These findings highlight a perception-production link mediated by positional cue reliance.

Prosody-modulated and vowel-dependent nasal merger in Taiwan Mandarin.

While Taiwan Mandarin alveolar and velar nasal codas are reported to lose their place contrast in a vowel-dependent fashion (i.e., more pervasive merging in the /iN/ context, followed by the /əN/ context, and no merging in the /aN/ context), other acoustic cues, such as vowel nasalization, have been reported to preserve the contrast. Results from our ultrasound experiment, in which Taiwan Mandarin speakers produced nasal codas in both focused and unfocused conditions, showed that the place contrast was most vulnerable to merger in /iN/ and /əN/ contexts and was not enhanced by focus. Instead, pre-nasal vowel nasalization emerged as a contrastive cue, especially in the highly merged /iN/ context in both focus conditions, suggesting a possible shift from place contrast to nasality contrast in production. A follow-up identification experiment revealed that speakers did not yet utilize the nasalization cues reliably to perceptually identify nasal codas, indicating a mismatch in the perception-production link.

Raw acoustic vs. normalized phonetic convergence: Imitation of the Northern Cities Shift in the American Midwest

Word shadowing tasks elicit phonetic convergence to the stimulus model talkers, suggesting a tight perception-production link. The magnitude of this convergence is affected by linguistic and social factors, suggesting that the perception-production link is mediated by higher-level phonological and social structures. The current study explored the nature of the perception-production link in an explicit comparison of raw acoustic vs. normalized phonetic convergence in word shadowing. American Midwestern participants repeated words after a model talker with features of the Northern Cities Shift vowels in one of three instruction conditions, which varied in whether participants were primed with the regional background of the model talker and in whether they were asked explicitly to imitate her. The results revealed normalized phonetic convergence to the model talker’s Northern Cities Shift vowels, even when this convergence entailed divergence from the raw acoustics, and token-by-token variability in her productions, consistent with a tight perception-production link that is mediated by linguistic structure. Modest effects of instruction condition on the magnitude of phonetic convergence were also observed, consistent with social information mediating this perception-production link. The results of this study provide converging evidence for phonetic convergence that is both phonetically-detailed and subject to constraint by higher-level representations.

Interpreting acoustic measures of imitation in word shadowing

Word shadowing tasks, in which a participant repeats words aloud after a pre-recorded model talker in a laboratory setting, elicit imitation of the model talker’s speech by the shadowing participant, even in the absence of explicit instructions to imitate. However, the nature of this imitation varies substantially within and across shadowers, model talkers, and linguistic features. Recent research has considered this variation in imitation in word shadowing to address how acoustic measures of imitation in these tasks should be interpreted. The results demonstrate (1) imitation of phonetic and phonological structure in vowel shifts and intonation contours, in addition to direct imitation of acoustic formant frequencies and f0; (2) more imitation of non-stereotyped social variants than of stereotyped social variants; and (3) both convergence to overall production patterns and synchrony with token-by-token variability in the speech of the model talker. Together, these results suggest a strong speech perception-production link that operates in real time and is mediated by linguistic and social representations.

Does perceptual high variability phonetic training improve L2 speech production? A meta-analysis of perception-production connection

AbstractThis meta-analysis of 31 studies aimed to determine the effectiveness of perception-based high variability phonetic training (HVPT) for second language (L2) production learning and to identify learner-related and methodological variables that influence production gains. Based on independent effect sizes for 43 within-participant and 17 between-participant designs, small-to-medium effects of post-training improvement were found. The average production gains for trained items and untrained items were 10.50% and 4.50%, respectively. Neither strong support for long-term retention of production learning nor generalization to untrained stimuli was observed, however. Moderator analyses showed that post-training production gains were influenced by a number of factors related to learner profiles (age and learning context), training features (provision of phonetic information, training duration, and training time per session), and features of production tests (elicitation tasks, prompt modality, and outcome measures). The relationship between perception and production gains was negligible at the participant level, but was significant and moderate at the level of individual studies for post-training and retention data. These findings provide partial support for a perception-production link. This study makes several recommendations for future studies investigating the effects of HVPT on L2 speech production learning.

Modelling Mandarin tone perception-production link through critical perceptual cues.

Theoretical accounts posit a close link between speech perception and production, but empirical findings on this relationship are mixed. To explain this apparent contradiction, a proposed view is that a perception-production relationship should be established through the use of critical perceptual cues. This study examines this view by using Mandarin tones as a test case because the perceptual cues for Mandarin tones consist of perceptually critical pitch direction and noncritical pitch height cues. The defining features of critical and noncritical perceptual cues and the perception-production relationship of each cue for each tone were investigated. The perceptual stimuli in the perception experiment were created by varying one critical and one noncritical perceptual cue orthogonally. The cues for tones produced by the same group of native Mandarin participants were measured. This study found that the critical status of perceptual cues primarily influenced within-category and between-category perception for nearly all tones. Using cross-domain bidirectional statistical modelling, a perception-production link was found for the critical perceptual cue only. A stronger link was obtained when within-category and between-category perception data were included in the models as compared to using between-category perception data alone, suggesting a phonetically and phonologically driven perception-production relationship.

On the effects of task focus and processing level on the perception–production link in second-language speech learning

AbstractThis study presents a reanalysis of existing data to investigate whether a relationship between perception and production abilities regarding a challenging second-language (L2) phonological contrast is observable (a) when both modalities must rely on accessing stored lexical representations and (b) when there is an asymmetry in task focus between perception and production. In the original studies, German learners of English were tested on their mastery of the English /ɛ/-/æ/ contrast in an auditory lexical decision task with phonological substitutions, a word-reading task, and a segmentally focused imitation task. Results showed that accurate nonword rejection in the lexical decision task was predicted by the Euclidean distance between the two vowels in word reading but not in imitation. These results extend previous findings to lexical perception and production, highlight the influence of task focus on the degree of coupling between the two modalities, and may have important implications for pronunciation training methods.

Competing targets in English sibilant imitation

This study investigates imitation of English /s/ to determine whether speakers converge toward normalized or raw acoustic targets. Participants exposed to increased spectral mean (SM) raised SM, converging toward both the raw acoustics of the model talker (who had high baseline SM) and the pattern of increased SM. However, after exposure to decreased SM, direction of shift depended on participant baseline. All participants converged to the raw acoustic values of the model talker, increasing or decreasing their own SM accordingly. These results suggest imitation is not necessarily mediated by perceptual normalization to different talkers, and raw acoustics can be the target of phonetic imitation. This has theoretical implications for the perception-production link and methodological implications for analysis of convergence studies.

Evidence for raw acoustics as the target of phonetic imitation

Phonetic imitation (also called convergence or accommodation) occurs when talkers alter their production towards speech they hear, even in lab settings without explicit instruction to imitate. Though general evidence for imitation is robust, much of the existing work does not distinguish between convergence towards a linguistic target versus convergence towards an acoustic target. This study presents a direct test case of these hypotheses through spontaneous imitation of the English alveolar sibilant /s/ in a delayed shadowing task. Participants first produced /s/-initial words, then were exposed to model speech with either enhanced or reduced spectral mean (SM) on /s/, and finally produced /s/-initial words again post-exposure. The model talker had higher than average baseline SM, so raw acoustic values were higher than those of most participants. All participants exposed to the enhanced stimuli raised SM, converging towards both the raw acoustics of the model talker and the linguistic pattern of enhancement. However, participants exposed to the reduced stimuli also raised SM. This diverged from the linguistic pattern of reduction but converged towards the raw acoustics of the model talker, whose SM was still higher than participants' even when reduced. These results suggest raw acoustics can be the target of phonetic imitation. Possible interactions with phonological contrast, implications for the perception-production link, and methodological considerations for future imitation studies are discussed.

The perception-production link in two types of cross-language mismatches

Previous data on the perception-production link have allowed for a variety of analyses, from the traditional straight-forward correlation between the two competencies (Flege 1995) to more nuanced interpretations of the data (Shultz et al. 2012). Where the traditional approach suggests that better perception is linked to better production, recent work challenges this conclusion and reveals a complex relationship. Here, we compare native French speakers' perception and production of English diphthongs (/a⌢ʊ/, /e⌢ɪ/ and /a⌢ɪ/) and interdental fricatives (/θ/ and /ð/). These specific segments were chosen as they are absent from the French phonological system, they show identifiable realizations, and they behave differently from one another. The competence in perception and production of 21 native hexagonal French speakers were collected through a repetition task and an AXB matching task. Consistent with Flege (1995), we show that for any given segment, greater perception accuracy predicts greater production accuracy. However, at the speaker level, the relationship between the two competences is a complex one, as the link between perception and production is not apparent in our data. These results would confirm the traditional approach at the segment level, but individual variation might prevents us from seeing it at the speaker level.

Producing and perceiving socially structured coarticulation: Coarticulatory nasalization in Afrikaans

Most theories of phonetics assume a tight relation between production and perception, and recent years have also seen increasing evidence for such a relation at the level of the individual. For the most part, however, this evidence comes from socially homogeneous speech communities where the targeted pattern of variation is mostly socially neutral. What implications might socially structured phonetic variation in the speech community have for the perception-production link? If listeners can predict the phonetic patterns of a talker based on the talker’s actual or assumed identity, would they adjust their perceptual strategies accordingly, possibly weakening the link between their own production and perception patterns? This study reports the results of a pair of experiments that investigate the production and perception of coarticulatory vowel nasalization in Afrikaans, a language for which variation in coarticulatory nasalization is socially structured. Relying on nasal airflow measures, the production experiment showed that speakers of White Afrikaans produce more extensive coarticulatory nasalization than speakers of Kleurling Afrikaans. The perception experiment used an eye-tracking paradigm to assess listeners’ perceptual reliance on coarticulatory nasalization, and found (i) that Afrikaans speakers’ use of coarticulatory nasalization in production predicts their perceptual reliance on this information, (ii) that they rapidly adjust to the coarticulatory timing patterns in the speech of other speakers, but also (iii) that they do not adjust their perceptual reliance on coarticulation in response to the assumed identity of the speaker. The link between perception and production therefore persists, even in this situation of socially structured variation in coarticulatory timing.

The relationship between the perception and production of L2 and L3 rhotics in young multilinguals; an exploratory cross-linguistic study

ABSTRACT There is a scarcity of research into the relationship between speech perception and production from a multilingual perspective. The present exploratory study investigates longitudinally the acquisition of rhotic sounds in 24 adolescent learners who have the same language repertoires (English, German, Polish), with English being their L2, and German and Polish their L1 or L3. The design involved a forced-choice goodness task (for perception) and a delayed repetition task (for production) administered in both L2 and L3 at two testing times. The overall results show better performance in perception than production of L2/L3 rhotics at both testing points, although the relationship between the two domains was modulated by the L1 background and universal learnability of the target rhotics. A major effect of language proficiency was found, as overall a stronger link between the two modalities was found for the L2 than the L3 of the learner groups. However, detailed analyses demonstrate different types of perception-production link for individual learners over time.

Categories in speech perception and production: How closely are they related?

A large body of research shows that fine phonetic details are not only used by listeners in processing speech (e.g., McMurray et al., 2009), but these details also affect listeners’ subsequent speech productions (e.g., Goldinger, 1998), suggesting a robust link between speech perception and production. However, evidence for a direct perception-production link is mixed, and the relationship between categories in perception and production is largely unknown. Newman (2003) found correlation between a perceptual prototype and mean VOT production for /pa/, while there were no production-perception correlations for other stop consonants. Nielsen (2021) found that perceptual boundaries in VOT vary widely across speakers, but that there is no apparent link between categorical boundary in perception of voicing contrast ([p]-[b]) and production variables of isolated speech (e.g., mean VOT for /p/ or /b/, the center of gap between two categories, the distance between the two category means), confirming Bailey and Haggard (1973). The current study further explores the relationship between perception and production categories by examining categorical boundaries of English stops using minimal pairs (e.g., pear-bear), their perceptual prototypes through a goodness rating task, and production variables in connected and isolated speech.

Examining the link between speech perception and production through categorical boundaries

The link between speech perception and production is one of the central topics in speech perception theories. Newman (2003) examined the correlations between acoustic measures of listeners’ perceptual and production prototypes for a given speech category, and found significant correlations for stop consonants (VOT) and for voiceless fricatives (spectral peaks). Shultz et al. (2012) also investigated the possible link between perception and production by examining speakers’ relative cue weighting of VOT and f0 in word initial stops. Although their results showed no significant correlation between weighting of VOT in production and perception, they reported a trend toward a slightly positive correlation between production and perception for VOT values. The current study extends these findings and aims to explore the perception-production link by examining the relationship between the location of categorical boundaries (between /p/ and /b/) in perception and production. If mental representations for perception and production are directly linked, then speakers with higher VOT values in both categories in production would be expected to also show a higher categorical boundary in perception.

ADVANCING THE STATE OF THE ART IN L2 SPEECH PERCEPTION-PRODUCTION RESEARCH: REVISITING THEORETICAL ASSUMPTIONS AND METHODOLOGICAL PRACTICES

AbstractOne of the basic goals of second language (L2) speech research is to understand the perception-production link, or the relationship between L2 speech perception and L2 speech production. Although many studies have examined the link, they have done so with strikingly different conceptual foci and methods. Even studies that appear to use similar perception and production tasks often present nontrivial differences in task characteristics and implementation. This conceptual and methodological variation makes meaningful synthesis of perception-production findings difficult, and it also complicates the process of developing new perception-production models that specifically address how the link changes throughout L2 learning. In this study, we scrutinize theoretical and methodological issues in perception-production research and offer recommendations for advancing theory and practice in this domain. We focus on L2 sound learning because most work in the area has focused on segmental contrasts.

Individual differences in the link between perception and production and the mechanisms of phonetic imitation

ABSTRACTThis study investigates the relationship between speech perception and production using explicit phonetic imitation. We used manipulated natural vowel (head-had) stimuli varying in spectral quality and duration in both perception and production tasks to explore the perception-production link in a direct and controlled way. We examined (1) whether individual listeners’ perceptual cue weights are related to their patterns of phonetic imitation and (2) phonological and perceptual constraints underlying phonetic imitation. Results showed that better perceptual abilities (i.e. larger cue weights) were related to better imitation of vowel duration. Furthermore, imitation of vowel spectral quality was mediated by contrast maintenance while vowel duration was not. Overall, vowel duration was better imitated despite being the less important cue perceptually. These results suggest that speech perception and production are indeed linked at the individual level, and both linguistic and perceptual-cognitive factors play a role in this process.

On the nature of the perception-production link: Individual variability in English sibilant-vowel coarticulation

This study aims to elucidate the nature of the perception–production link with respect to coarticulation by examining the production and perception of English sibilants before different vowels. A group of native speakers of American English were recorded reciting a set of /s/- and /ʃ/-initial words in different vocalic contexts and took part in an identification experiment designed to test their ability to adjust their perceptual expectation in light of the vocalic influence on the preceding sibilant. Significant correlations between the production and perception results were observed when by-subject estimates for context-relevant predictors (and their interactions) in the perception regression models were examined in relation to the by-subject estimates of the production regression models. These results suggest a positive correlation between how much an individual attends to context-specific variation in perception and how the sibilant contrast is realized in specific vocalic contexts. Ramifications of these findings are discussed for the nature of speech perception and production and the understanding of sound change.

Imitation in a Second Language Relies on Phonological Categories but Does Not Reflect the Productive Usage of Difficult Sound Contrasts.

This study investigated the relationship between imitation and both the perception and production abilities of second language (L2) learners for two non-native contrasts differing in their expected degree of difficulty. German learners of English were tested on perceptual categorization, imitation and a word reading task for the difficult English /ɛ/-/æ/ contrast, which tends not to be well encoded in the learners' phonological inventories, and the easy, near-native /i/-/ɪ/ contrast. As expected, within-task comparisons between contrasts revealed more robust perception and better differentiation during production for /i/-/ɪ/ than /ɛ/-/æ/. Imitation also followed this pattern, suggesting that imitation is modulated by the phonological encoding of L2 categories. Moreover, learners' ability to imitate /ɛ/ and /æ/ was related to their perception of that contrast, confirming a tight perception-production link at the phonological level for difficult L2 sound contrasts. However, no relationship was observed between acoustic measures for imitated and read-aloud tokens of /ɛ/ and /æ/. This dissociation is mostly attributed to the influence of inaccurate non-native lexical representations in the word reading task. We conclude that imitation is strongly related to the phonological representation of L2 sound contrasts, but does not need to reflect the learners' productive usage of such non-native distinctions.

Examining the Temporal Structure of the Perception–Production Link in Second Language Acquisition: A Longitudinal Study

AbstractMost studies on the perception–production link have assumed a synchronous relationship according to which gains in perception transfer to production rapidly and efficiently. However, time‐lagged and asymptotic relationships are also possible, where perception would guide production at a later stage or production would improve only once perception has reached a high level of accuracy. This study investigated the temporal dynamics of the perception–production link by modeling English speakers’ ability to perceive and produce second language (L2) Spanish stops over time. Mixed‐effects modeling of the d′ perception and voice onset time (VOT) production data demonstrated significant development in both areas. Time‐lagged change models indicated a change in d′ significantly related to decreasing VOT in L2 /p/ at the following testing time, but no significant relationships emerged between perception and production of L2 /b/. Multiple patterns among individuals emerged. Results are interpreted in light of contemporary models of L2 speech learning.Open PracticesThis article has been awarded an Open Materials badge. Study materials are publicly accessible in the IRIS digital repository at http://www.iris-database.org. Learn more about the Open Practices badges from the Center for Open Science: https://osf.io/tvyxz/wiki.

The function and mechanism of vocal accommodation in humans and other primates.

The study of non-human animals, in particular primates, can provide essential insights into language evolution. A critical element of language is vocal production learning, i.e. learning how to produce calls. In contrast to other lineages such as songbirds, vocal production learning of completely new signals is strikingly rare in non-human primates. An increasing body of research, however, suggests that various species of non-human primates engage in vocal accommodation and adjust the structure of their calls in response to environmental noise or conspecific vocalizations. To date it is unclear what role vocal accommodation may have played in language evolution, in particular because it summarizes a variety of heterogeneous phenomena which are potentially achieved by different mechanisms. In contrast to non-human primates, accommodation research in humans has a long tradition in psychology and linguistics. Based on theoretical models from these research traditions, we provide a new framework which allows comparing instances of accommodation across species, and studying them according to their underlying mechanism and ultimate biological function. We found that at the mechanistic level, many cases of accommodation can be explained with an automatic perception-production link, but some instances arguably require higher levels of vocal control. Functionally, both human and non-human primates use social accommodation to signal social closeness or social distance to a partner or social group. Together, this indicates that not only some vocal control, but also the communicative function of vocal accommodation to signal social closeness and distance must have evolved prior to the emergence of language, rather than being the result of it. Vocal accommodation as found in other primates has thus endowed our ancestors with pre-adaptations that may have paved the way for language evolution.