Context Effects In Speech Perception Research Articles

What do voices contribute to acoustic context effects in speech perception?

All perception takes place in context, and speech perception is no exception. Surrounding sounds form a context that guides speech perception. This is particularly true when earlier sounds inform the perception and/or recognition of later sounds. Several lines of research have advanced models where listeners use preceding context to account for characteristics of the talker’s voice, such as vocal tract properties, articulatory maneuvers, or the speaker’s identity. Conversely, research with nonspeech sounds and/or nonhuman animals has advanced alternative models where perception is responding more to signal acoustics than to vocal characteristics. In this talk, I will review recent evidence from our research on acoustic context effects and talker normalization when hearing speech and nonspeech (music) sounds. Parallel patterns of performance across speech and music domains are conducive to, but not definitive evidence of, acoustic context being utilized according to general signal acoustics, with speech- and voice-specific contributions coming later in the processing stream.

Internal Cognitive Load Differentially Influences Acoustic and Lexical Context Effects in Speech Perception: Evidence From a Population With Attention-Deficit/Hyperactivity Disorder.

To overcome variability in spoken language, listeners utilize various types of context information for disambiguating speech sounds. Context effects have been shown to be affected by cognitive load. However, previous results are mixed regarding the influence of cognitive load on the use of context information in speech perception. We tested a population characterized by an attention-deficit/hyperactivity disorder (ADHD) to better understand the relationship between attention (or internal cognitive load) and context effects. The use of acoustic versus lexical properties of the surrounding signal to disambiguate speech sounds was examined in listeners with ADHD and neurotypical listeners. Compared to neurotypicals, individuals with ADHD relied more strongly on lexical context for speech perception; however, reliance on acoustic context information from speech rate did not differ. These findings confirm that cognitive load impacts the use of high-level but not low-level context information in speech and imply that speech recognition deficits in ADHD likely arise due to impaired higher order cognitive processes.

Evaluating peripheral versus central contributions to spectral context effects in speech perception

Perception of a sound is influenced by spectral properties of surrounding sounds. When frequencies are absent in a preceding acoustic context before being introduced in a subsequent target sound, detection of those frequencies is facilitated via an auditory enhancement effect (EE). When spectral composition differs across a preceding context and subsequent target sound, those differences are perceptually magnified and perception shifts via a spectral contrast effect (SCE). Each effect is thought to receive contributions from peripheral and central neural processing, but the relative contributions are unclear. The present experiments manipulated ear of presentation to elucidate the degrees to which peripheral and central processes contributed to each effect in speech perception. In Experiment 1, EE and SCE magnitudes in consonant categorization were substantially diminished through contralateral presentation of contexts and targets compared to ipsilateral or bilateral presentations. In Experiment 2, spectrally complementary contexts were presented dichotically followed by the target in only one ear. This arrangement was predicted to produce context effects peripherally and cancel them centrally, but the competing contralateral context minimally decreased effect magnitudes. Results confirm peripheral and central contributions to EEs and SCEs in speech perception, but both effects appear to be primarily due to peripheral processing.

Acoustic context effects in speech perception.

The extreme acoustic variability of speech is well established, which makes the proficiency of human speech perception all the more impressive. Speech perception, like perception in any modality, is relative to context, and this provides a means to normalize the acoustic variability in the speech signal. Acoustic context effects in speech perception have been widely documented, but a clear understanding of how these effects relate to each other across stimuli, timescales, and acoustic domains is lacking. Here we review the influences that spectral context, temporal context, and spectrotemporal context have on speech perception. Studies are organized in terms of whether the context precedes the target (forward effects) or follows it (backward effects), and whether the context is adjacent to the target (proximal) or temporally removed from it (distal). Special cases where proximal and distal contexts have competing influences on perception are also considered. Across studies, a common theme emerges: acoustic differences between contexts and targets are perceptually magnified, producing contrast effects that facilitate perception of target sounds and words. This indicates enhanced sensitivity to changes in the acoustic environment, which maximizes the amount of potential information that can be transmitted to the perceiver. This article is categorized under: Linguistics > Language in Mind and Brain Psychology > Perception and Psychophysics.

Natural speech statistics shift phoneme categorization.

All perception takes place in context. Recognition of a given speech sound is influenced by the acoustic properties of surrounding sounds. When the spectral composition of earlier (context) sounds (e.g., more energy at lower first formant [F1] frequencies) differs from that of a later (target) sound (e.g., vowel with intermediate F1), the auditory system magnifies this difference, biasing target categorization (e.g., towards higher-F1 /ɛ/). Historically, these studies used filters to force context sounds to possess desired spectral compositions. This approach is agnostic to the natural signal statistics of speech (inherent spectral compositions without any additional manipulations). The auditory system is thought to be attuned to such stimulus statistics, but this has gone untested. Here, vowel categorization was measured following unfiltered (already possessing the desired spectral composition) or filtered sentences (to match spectral characteristics of unfiltered sentences). Vowel categorization was biased in both cases, with larger biases as the spectral prominences in context sentences increased. This confirms sensitivity to natural signal statistics, extending spectral context effects in speech perception to more naturalistic listening conditions. Importantly, categorization biases were smaller and more variable following unfiltered sentences, raising important questions about how faithfully experiments using filtered contexts model everyday speech perception.

Variability in talkers' fundamental frequencies shapes context effects in speech perception.

The perception of any given sound is influenced by surrounding sounds. When successive sounds differ in their spectral compositions, these differences may be perceptually magnified, resulting in spectral contrast effects (SCEs). For example, listeners are more likely to perceive /ɪ/ (low F1) following sentences with higher F1 frequencies; listeners are also more likely to perceive /ɛ/ (high F1) following sentences with lower F1 frequencies. Previous research showed that SCEs for vowel categorization were attenuated when sentence contexts were spoken by different talkers [Assgari and Stilp. (2015). J. Acoust. Soc. Am. 138(5), 3023-3032], but the locus of this diminished contextual influence was not specified. Here, three experiments examined implications of variable talker acoustics for SCEs in the categorization of /ɪ/ and /ɛ/. The results showed that SCEs were smaller when the mean fundamental frequency (f0) of context sentences was highly variable across talkers compared to when mean f0 was more consistent, even when talker gender was held constant. In contrast, SCE magnitudes were not influenced by variability in mean F1. These findings suggest that talker variability attenuates SCEs due to diminished consistency of f0 as a contextual influence. Connections between these results and talker normalization are considered.

Lexical tone is perceived relative to locally surrounding context, vowel quality to preceding context.

Important speech cues such as lexical tone and vowel quality are perceptually contrasted to the distribution of those same cues in surrounding contexts. However, it is unclear whether preceding and following contexts have similar influences, and to what extent those influences are modulated by the auditory history of previous trials. To investigate this, Cantonese participants labeled sounds from (a) a tone continuum (mid- to high-level), presented with a context that had raised or lowered fundamental frequency (F0) values and (b) a vowel quality continuum (/u/ to /o/), where the context had raised or lowered first formant (F1) values. Contexts with high or low F0/F1 were presented in separate blocks or intermixed in 1 block. Contexts were presented following (Experiment 1) or preceding the target continuum (Experiment 2). Contrastive effects were found for both tone and vowel quality (e.g., decreased F0 values in contexts lead to more high tone target judgments and vice versa). Importantly, however, lexical tone was only influenced by F0 in immediately preceding and following contexts. Vowel quality was only influenced by the F1 in preceding contexts, but this extended to contexts from preceding trials. Contextual influences on tone and vowel quality are qualitatively different, which has important implications for understanding the mechanism of context effects in speech perception. (PsycINFO Database Record

Background noise interacts with spectral context effects during speech categorization

Speech perception is heavily influenced by surrounding sounds. When spectral properties differ between earlier (context) and later (target) sounds, this can produce spectral contrast effects (SCEs) that bias categorization of later sounds. For example, context sounds with a low-F1 bias produce more high-F1 responses to the target vowel and vice versa. SCEs have been studied for decades but only for speech perception in quiet. Background noise is expected to interfere with these effects, but the extent of this interference is unclear. On each trial, listeners heard a context sentence and target vowel both combined with speech-shaped noise. Sentences were filtered to add a + 20-dB spectral peak at low-F1 (100–400 Hz) or high-F1 frequencies (550–850 Hz). Target vowels varied from /ɪ/ to /ɛ/. Background noise was unmodulated or sinusoidally modulated at 2/4/8 Hz at variable SNRs. In general, SCEs were larger at better SNRs and in unmodulated noise, but all SCEs were smaller than those reported for speech in quiet. Complex patterns of results were observed at different noise modulation rates, with likely ties to modulations in F1 frequency regions in the context sentence. In sum, background noise diminishes context effects in speech perception, but the extent depends on noise characteristics.

ERP signatures of cross-modal semantic fragment priming: Early context effects in speech perception

Research in speech perception and language comprehension has shown that the processing and recognition of a word (the target) is facilitated if a related word (the prime) precedes it. This effect is observable even if the prime only consists of a word-initial fragment (e.g. bor- facilitates the recognition of border). The fragment priming paradigm has gained an important role in ERP studies on lexical access, since it reliably shows a distinct deflection at around 350 ms after word-onset, taken as index for lexical activation. While previous studies concentrated on form-related prime–target pairs, we are interested in ERP fragment priming components elicited by meaning-related pairs. For that purpose, we designed a cross-modal priming experiment, in which primes were either the initial syllables of their following targets (e.g. bor[ der]– border; phonological condition) or the initial syllables of a semantic relative to these targets (e.g. mar[ gin]– border; semantic condition). We found a reliable P350 effect in the phonological condition, but not in the semantic condition, and a N400 priming effect for the phonological condition. The phonological and the semantic condition differed from the control condition in an early negativity around 180 ms post onset. Our results are discussed with respect to ERP components that index form-based and meaning-based processing, possibly reflecting separate representations.

Presence of preceding sound affects the neural representation of speech sounds: Frequency following response data.

A substantial body of literature has focused on context effects in speech perception in which manipulation of the phonemic or spectral content of preceding sounds (e.g., /al/ versus /ar/) result in a shift in the perceptual categorization of a target syllable (e.g., /da/ versus /ga/). In a previous study utilizing the frequency‐following response (FFR) to measure neural correlates of these context effects [R. Aravamudhan, J. Acoust. Soc. Am. 126, 2204], it was noted that the representation of target formant trajectories were much weaker when the stimulus was presented in isolation versus following some type of context. To examine this effect explicitly, a series of syllables varying from /da/ to /ga/ was presented to listeners either in isolation or following the syllables /a/, /al/, or /ar/ (with a 50‐ms silent gap between context and target). FFR measures were obtained from EEG recordings while participants listened passively. The resulting narrow‐band spectrograms over the grand averages demonstrated that the target formant representations were much weaker or absent when presented in isolation. This counter‐intuitive result may indicate the importance of context for robust encoding of important spectral features in speech and provide a cautionary note on the use of isolated syllables in auditory testing.

Effect of preceding speech on nonspeech sound perception

Data from Japanese quail suggest that the effect of preceding liquids (/l/ or /r/) on response to subsequent stops (/g/ or /d/) arises from general auditory processes sensitive to the spectral structure of sound [A. J. Lotto, K. R. Kluender, and L. L. Holt, J. Acoust. Soc. Am. 102, 1134–1140 (1997)]. If spectral content is key, appropriate nonspeech sounds should influence perception of speech sounds and vice versa. The former effect has been demonstrated [A. J. Lotto and K. R. Kluender, Percept. Psychophys. 60, 602–619 (1998)]. The current experiment investigated the influence of speech on the perception of nonspeech sounds. Nonspeech stimuli were 80-ms chirps modeled after the F2 and F3 transitions in /ga/ and /da/. F3 onset was increased in equal steps from 1800 Hz (/ga/ analog) to 2700 Hz (/da/ analog) to create a ten-member series. During AX discrimination trials, listeners heard chirps that were three steps apart on the series. Each chirp was preceded by a synthesized /al/ or /ar/. Results showed context effects predicted from differences in spectral content between the syllables and chirps. These results are consistent with the hypothesis that spectral contrast influences context effects in speech perception. [Work supported by ONR, NOHR, and CNBC.]

Neighboring spectral content influences vowel identification.

Four experiments explored the relative contributions of spectral content and phonetic labeling in effects of context on vowel perception. Two 10-step series of CVC syllables ([bVb] and [dVd]) varying acoustically in F2 midpoint frequency and varying perceptually in vowel height from [delta] to [epsilon] were synthesized. In a forced-choice identification task, listeners more often labeled vowels as [delta] in [dVd] context than in [bVb] context. To examine whether spectral content predicts this effect, nonspeech-speech hybrid series were created by appending 70-ms sine-wave glides following the trajectory of CVC F2's to 60-ms members of a steady-state vowel series varying in F2 frequency. In addition, a second hybrid series was created by appending constant-frequency sine-wave tones equivalent in frequency to CVC F2 onset/offset frequencies. Vowels flanked by frequency-modulated glides or steady-state tones modeling [dVd] were more often labeled as [delta] than were the same vowels surrounded by nonspeech modeling [bVb]. These results suggest that spectral content is important in understanding vowel context effects. A final experiment tested whether spectral content can modulate vowel perception when phonetic labeling remains intact. Voiceless consonants, with lower-amplitude more-diffuse spectra, were found to exert less of an influence on vowel perception than do their voiced counterparts. The data are discussed in terms of a general perceptual account of context effects in speech perception.

Quantification of context effects in speech perception: Influence of prosody

Speech recognition requires that the auditory system incorporate both acoustic and contextual cues from the signal for adequate comprehension. A number of contextual effects in speech have been shown to improve recognition ability, among these are intonational/prosodic cues and syntactic and semantic congruency. The relationship between the processing of prosodic cues and the processing of syntactic and semantic information, however, is uncertain. The present investigation mathematically quantified the contribution of linguistic factors to speech perception, with an emphasis on relating how prosody affects the processing of syntactic and semantic information. Subjects listened to and repeated sentences having normal and abnormal prosodic contours, which contained three levels of syntactic and semantic constraint. Results revealed that prosody does affect listeners' ability to utilize syntactic and semantic information, suggesting that the presence of prosody serves to alter how an individual utilizes context. Such findings indicate that prosody may be a more important cue for speech perception, with listeners placing greater weight on the prosodic redundancies in the speech signal than on the syntactic and semantic redundancies when processing speech. Implications of these data are discussed.

Auditory‐visual context effects on the perception of /r/ and /l/ in a stop cluster.

Context effects in speech perception are thought to reflect knowledge about coarticulatory influences from the surrounding phonetic environment. This study investigated if such effects occur when the context is specified in the visual modality and segmental information in the auditory modality. In the first experiment, two continua were synthesized: one varying from /iri–ili/ and the other from /ibri–ibli/. When presented to listeners for identification as /r/ or /l/, there was a significant shift in the category boundary between the two continua. In a second experiment, the /iri–ili/ tokens were paired with visual tokens of a talker saying /ibi/. When presented in an auditory‐visual (AV) condition, these tokens were perceived as ranging from /ibri/ to /ibli/. Observers also identified the /iri–ili/ tokens in a separate auditory‐only (AO) condition. Results indicated a similar shift in the /r–l/ boundary between the AO and AV conditions. Analysis of /r/ and /l/ productions revealed that the perceptual adjustments were consistent with the acoustic consequences of articulating /r/ and /l/ in a stop cluster. The findings suggest that the perceptual adjustment reflects cross‐modal knowledge of the coarticulatory effects of a bilabial consonant on the acoustic realization of /r/ and /l/. [Work supported by NIDCD, NIH.]

Context effects in a double-weak theory of speech perception.

The present study provides an elaboration of the "double-weak" theory of speech perception proposed by Nearey (1990, 1991). In this framework, the objects of speech perception (and production) are viewed as neither primarily auditory nor primarily gestural; rather, they are abstract, symbolic elements lawfully constrained to map onto relatively simple (but not entirely transparent) patterns in both domains. Speech cannot be understood unless both articulation and acoustics are considered: Many production strategies appear to be directed at achieving acoustically-oriented goals, yet most context effects in speech perception seem to be motivated by the consequences of gestural overlap. The double-weak framework suggests that speech perception and speech production are less-than-perfect inverses of each other. Despite long-term accommodation of each for the demands of the other, real-time production and perception may operate as autonomous subsystems. A family of perceptual models is discussed that provides varying degrees of approximation to "ideal solutions" (in the sense of minimizing error rate) to classifying production data exhibiting contextual interactions. Members of this family that provide substantial, yet incomplete, compensation for the consequences of gestural overlap appear to be adequate to account for the results of many speech perception experiments. Such partial perceptual compensation allows, in principle, for the kind of imperfect "error correction" discussed by Ohala (1981, 1990) in conjunction with hypo- and hypercorrection phenomena.

Macaques show context effects in speech perception

The phonetic distinction between the stop consonant /b/ and semivowel /w/ is cued primarily by duration of the initial formant transition. Syllable duration, however, is a context that influences human adults' and infants' perception of this distinction. For example, when overall syllable duration is long (296 ms), infants show poor discrimination of syllables with transition durations of 16 and 40 ms, but good discrimination of syllables with transition durations of 40 and 64 ms. However, when overall syllable duration is short (80 ms), the opposite is true [P. D. Eimas and J. L. Miller, Science 209, 1140–1141 (1980)]. The current study addresses whether rhesus and Japanese macaques' perception of the /b-w/ distinction is also influenced by syllable duration: Six CV syllables used in the Eimas and Miller (1980) study were presented to five macaques using a same-different discrimination paradigm. Results showed that macaques' pattern of discrimination was similar to the pattern shown by infants, thus supporting an auditory-based mechanism for the effect of syllable duration in phonetic perception. [Work supported by NSF and NIH.]