Abstract
The goal of this developmental speech perception study was to assess whether and how age group modulated the influences of high-level semantic context and low-level fundamental frequency (F0) contours on the recognition of Mandarin speech by elementary and middle-school-aged children in quiet and interference backgrounds. The results revealed different patterns for semantic and F0 information. One the one hand, age group modulated significantly the use of F0 contours, indicating that elementary school children relied more on natural F0 contours than middle school children during Mandarin speech recognition. On the other hand, there was no significant modulation effect of age group on semantic context, indicating that children of both age groups used semantic context to assist speech recognition to a similar extent. Furthermore, the significant modulation effect of age group on the interaction between F0 contours and semantic context revealed that younger children could not make better use of semantic context in recognizing speech with flat F0 contours compared with natural F0 contours, while older children could benefit from semantic context even when natural F0 contours were altered, thus confirming the important role of F0 contours in Mandarin speech recognition by elementary school children. The developmental changes in the effects of high-level semantic and low-level F0 information on speech recognition might reflect the differences in auditory and cognitive resources associated with processing of the two types of information in speech perception.
Introduction
Speech recognition involves smooth and rapid integration of perceptual and cognitive systems, during which lower- and higher-level information sources in the spoken language are analyzed and synthesized (). In adverse listening conditions, acoustic-phonetic cues are often insufficient for the identification of particular words because of speech signal quality (e.g., distortion to speech sounds caused by poor articulation) or listening background (e.g., the presence of noise or interfering speech). In such circumstances, various supportive mechanisms, especially the semantic context from the preceding words in an utterance or discourse can help the listener perceive, recognize, and understand subsequent words. Previous studies have shown that these factors including quality of speech signal, listening background, and semantic context interact with each other in jointly affecting speech recognition outcome (; ; ). For example, found that acoustically degraded sentences with flattened fundamental frequency (the lowest frequency of a periodic complex sound and is the acoustic correlate of pitch for speech and music, hereafter referred to as F0) contours were as intelligible as original sentences with natural F0 contours when both sentence types were presented in quiet. However, the flattened F0 contours yielded decreased speech intelligibility in white noise and multi-talker babble backgrounds, indicating the significant interaction between the quality of speech signal and listening background during speech recognition. showed that both older and young adults derived greater contextual benefit for recognizing the sentence-final word in the presence of masking broadband noise when compared with quiet, indicating the significant interaction between listening condition and semantic context for speech intelligibility.
The task of listening to speech under suboptimal circumstances is challenging for listeners of all ages, but the detrimental effects of degraded speech and auditory masking appear to be particularly large during childhood because children are not born with fully mature auditory processing capabilities but instead demonstrate a developmental time course of improvement in processing complex auditory situations (; ; ; ; ). Children’s difficulty in perceiving speech in noise and against interfering speech has been investigated in a number of previous studies, which have consistently shown that children need more favorable conditions than adults to achieve comparable speech-in-noise scores (; ; ). For example, examined sentence recognition by young adults and children of three age groups (5, 9, and 11 years) in the presence of an eight-talker babble and found that children of all the three groups required a more favorable signal-to-noise ratio (SNR) to perform as well as the young adults. This child–adult difference was largest for the youngest children, reflecting the developmental progress in the ability to recognize speech in multi-talker babbles. suggested a similar disadvantage for children in speech recognition in noise when comparing the performances of 6- to 15-year-old children and young adults on the NU-CHIPS (the Northwestern University Children’s Perception of Speech test) in quiet, steady-state and interrupted background noise. The younger children exhibited greater difficulty for listening in both kinds of background noise than the older children and adult-like performance was only exhibited in the children 11 years of age and older.
Previous work suggests that younger children are able to use semantic context to assist speech recognition in quiet as early as 2-years-old involving predictive coding to recognize the sentence-final nouns (). The ability to use semantic context in adverse listening backgrounds continues to improve into adolescence with significant interactions among the factors of listening background, semantic context and age. For example, investigation of speech intelligibility of high- and low-predictability sentences in children at 9–17 years of age showed that contextual cues assisted children of all ages in identifying words masked by babbles with older children performing better than younger children in the high context conditions (). Similar studies including children aged 5–11 years () or 5–10 years () also supported age-related improvement in contextual benefit for recognizing speech in adverse listening backgrounds.
In addition to the line of work on the effects of adverse listening backgrounds, there has been another line of investigation on the effect of degraded speech signal on speech recognition by children. found that clear speech with more pauses, longer sentence/vowel durations and greater vowel spaces was more intelligible than conversational speech for children aged 6–13 years. In addition, children have the ability to use semantic context to decode degraded signals in quiet. For example, in a quiet listening condition, found that 4-years-old were better at detecting mispronounced words in high-predictability sentences than in low-predictability sentences. found that children 8–10 years of age required shorter portions of the target words (i.e., less acoustic-phonetic information) for recognition in high-predictability than in low-predictability sentences. By contrast, children aged 5–7 years required as much acoustic-phonetic information with high-predictability as with low-predictability contexts to identify target words.
To our knowledge, previous studies have not combined the two lines of work to investigate whether children can use sentence context to aid recognition of degraded speech in adverse listening backgrounds. There is also a lack of studies in which the primary target language is not English to test the generalizability of previous findings on other languages with typologically distinct acoustic and prosodic properties. In a tonal language like Mandarin Chinese, F0 is primarily used to distinguish lexical meanings (). This differs greatly from non-tonal languages (e.g., English and French) in which F0 information is mainly used for pragmatic purposes such as sentence modality (interrogative vs. declarative), emphasis and emotion (). Although lexical tones have a phonemic role in Mandarin, alteration of the original F0 contours has little effect on the intelligibility of sentences in quiet. In contrast, adverse listening conditions (e.g., interfering speech) substantially reduce the intelligibility of Mandarin sentences with altered F0 contours compared with sentences with natural F0 patterns (; ). Developmental research reveals that Mandarin-speaking children show sensitivity to F0 information several days after birth () and acquire the F0 patterns of lexical tones in perception and production at an early age (). Considering the interaction of sentential semantic context, F0 contours and listening condition on Mandarin speech recognition by young adults (), it remains unclear how these factors may separately or jointly affect a child’s performance.
The purpose of the present study aimed to investigate the possible different effects of high-level semantic and low-level F0 information on the recognition of Mandarin speech by elementary and middle school children in quiet and single-talker interference conditions. On the basis of previous studies that have suggested a developmental/maturational course for children’s ability to recognize speech in adverse listening conditions (; ; ), we predicted that the middle school-aged children would perform better than the elementary school children in the presence of interfering speech, while the two groups would perform equally well in quiet. Furthermore, the middle school-aged children would benefit more from sentence context than the elementary school children in the interfering background but not in quiet. More importantly, we expected a significant interaction between age group and F0 contours. Specifically, the middle school children would perform better than the elementary school children when F0 contours were flattened, while both groups would perform similarly on recognizing speech with natural F0 contours, assuming that younger children might rely more on natural F0 patterns than older children in speech recognition. The effect of age group on F0 contours might be further modulated by listening background and semantic context, e.g., the middle school children could benefit more from semantic context than the elementary school children when F0 patterns changed from natural contours to flat contours.
Materials and Methods
Subjects
Forty-eight elementary school children (23 boys, Grade 4, age range 9.6–10.1 years) and forty-six middle school children (25 boys, Grade 8, age range 13.5–14.2 years) from local schools in Beijing were recruited. All children were native speakers of Mandarin Chinese and none of the children had a history of any neurological or psychiatric disorders. Audiometric evaluation was carried out using a GSI 61 clinical audiometer (Grason-Stadler, Inc., Madison, WI, United States) with a standard Hughson-Westlake approach and each child had pure-tone thresholds no greater than 20 dB HL at octave intervals from 250 to 8000 Hz bilaterally. Written informed consent was obtained from the children and their legal guardians. The study was approved by the Institutional Review Boards (IRBs) at Beijing Language and Culture University and Beijing Normal University, and was conducted in compliance with the Declaration of Helsinki.
This study focused on the two age groups because children 11 years of age and older exhibited adult-like performance on speech recognition in noise (). A mixed-design was adopted with semantic context (normal sentence vs. word list sentence) and F0 contours (normal vs. flat) as within-subject factors, while listening background (quiet vs. interference) and age group (elementary school vs. middle school) as between-subject factors. The participants (boys/girls) randomly divided in the 4 between-subject conditions were distributed as follows: elementary school/quiet, number of subjects (n) = 24 including 11 boys and 13 girls; elementary school/interference background, n = 24 including 12 boys and 12 girls; middle school/quiet, n = 23 including 13 boys and 10 girls; middle school/interference background, n = 23 including 12 boys and 11 girls.
Stimuli
Four types of sentences were used as targets. Specifically, the target sentences included normal/word list sentences with natural/flat F0 contours, respectively, manipulating semantic context and F0 contours. Such manipulations have been adopted in our previous studies (; ; ). The normal sentences were 28 declarative Chinese sentences with each sentence comprised of 3 to 6 words (2–4 content words plus 0–2 functional words) that were familiar to both the elementary and middle school children. Words from the entire pool of the normal sentences were pseudo-randomly selected to form the word list sentences matched in number of content and functional words with the normal sentences. Word list sentences were syntactically correct but semantically meaningless at the whole sentence level. That is, recognition of both word list and normal sentence recruits syntactic and semantic processing, but only normal sentence provides the meaningful semantic context of “who does what to whom” beyond random lexical semantics. A male native Chinese speaker read the normal and word list sentences. Monotonous sentences with flat F0 contours were created by flattening the natural F0 contours at each sentence’s mean F0 (Figure 1).
FIGURE 1
Masker stimuli were consonant-misplaced sentences, which were constructed based on the normal sentences. Specifically, the initial consonant of every syllable in each normal sentence was replaced with another consonant as long as the phonotactic rules of Mandarin were not violated. Consonant-misplaced sentences were unintelligible at both word and sentence levels, thus having minimal effect of informational masking (; ). A female native Mandarin speaker read the masker sentences, which was to enable separation of the target message from the interfering speech. Without this indexical information, participants would have to be trained on the identity of the target speaker. The listeners were instructed to listen to the male speaker for the content rather than the identity of the target speaker throughout (). The target and masker sentences were first edited to be at the sound pressure levels of 75 and 70 dB, respectively, and then combined to form the speech against interference stimuli with SNR level set at +5 dB.
Procedures
We followed similar experimental procedures in our previous studies (; ), except that listeners were instructed to verbally repeat rather than write down the whole target normal/word list sentences in order to reduce cognitive burden on the children. Their responses were recorded and scored by the first author of this paper and checked by an independent auditor blind to the experiment. A strict score standard was adopted. Specifically, only words with consonants, vowels and lexical tones all correctly identified were considered correct answers. During the experiment, the listeners were seated in a sufficiently silent room (not higher than 15 dB for ambient noise level). A set of Edifier R18 loudspeakers were used to present the stimuli with the sound level set at 65 dB SPL calibrated at the listeners’ head. Because F0 contours and semantic context were within-subject factors, each listener was presented with a total of 56 trials—14 normal sentences and 14 word list sentences with natural or flat F0 contours. Each child listened to half of the total normal sentences/wordlist sentences (14/28) with natural F0 contours and the other half with flat F0 contours, i.e., each child did not listen to the same normal sentence/wordlist sentence with natural or flat F0 contours. F0 patterns were counterbalanced across subjects. The experiment adopted self-paced paradigm and the children were encouraged to guess which words they heard. Each normal/word list sentence was used only once with each subject. Before the actual experiment, all children participated in a brief practice session representing samples of all conditions.
Results
A keyword-correct count was used to calculate the speech recognition sore, i.e., percentage of content words correctly identified (; ) (Table 1). In order to test the main and interaction effects, a repeated measures ANOVA was conducted with listening condition and age group as between-subject factors, and sentence context and F0 contours as within-subject factors. Results revealed significant main effects of all the four factors [semantic context: F(1,90) = 275.877, p < 0.001, = 0.754; listening background: F(1,90) = 112.168, p < 0.001, = 0.555; F0 contours: F(1,90) = 384.19, p < 0.001, = 0.81; age group, F(1,90) = 32.236, p < 0.001, = 0.264], and four significant two-way interaction effects as well [interaction between semantic context and listening condition: F(1,90) = 12.53, p < 0.001, = 0.122; interaction between F0 contours and age group: F(1,90) = 22.464, p < 0.001, = 0.2; interaction between listening background and F0 contours: F(1,90) = 32.48, p < 0.001, = 0.265; and interaction between F0 contours and semantic context: F(1, 90) = 4.866, p = 0.03, = 0.051].
Table 1
| Natural F0 | Flat F0 | |||||
|---|---|---|---|---|---|---|
| Normal sentence | Word list sentence | Normal sentence | Word list sentence | |||
| Elementary school children | ||||||
| Quiet | 0.94 (±0.05) | 0.84 (±0.08) | 0.82 (±0.12) | 0.66 (±0.13) | ||
| +5 dB SNR | 0.83 (±0.12) | 0.63 (±0.12) | 0.54 (±0.14) | 0.39 (±0.16) | ||
| Middle school children | ||||||
| Quiet | 0.97 (±0.05) | 0.90 (±0.05) | 0.92 (±0.04) | 0.77 (±0.10) | ||
| + 5 dB SNR | 0.89 (±0.08) | 0.72 (±0.13) | 0.75 (±0.14) | 0.54 (±0.13) | ||
Mean accuracy (±standard deviation) of each condition for both groups.
There were also two significant 3-way interaction effects [interaction between age group, F0 contours and semantic context: F(1,90) = 5.271, p = 0.024, = 0.055, and interaction between F0 contours, semantic context and listening background: F(1,90) = 7.292, p = 0.008, = 0.075]. The four-way interaction effect was not significant [F(1,90) = 1.155, p = 0.285, = 0.013]. Considering the primary purpose of the present study was to investigate the developmental difference between elementary and middle school children in the use of semantic context and F0 contours during Mandarin speech recognition, the following analyses focused on the significant 3-way interaction between age group, F0 contours and semantic context. Specifically, this interaction effect was decomposed step by step (). First, separate two-way ANOVAs were carried out to examine the simple interaction effects. Results showed that significant interaction between F0 contours and semantic context was observed in the middle school children [F(1,45) = 7.837, p = 0.008, = 0.151], but absent in the elementary school children [F(1,47) = 0.006, p = 0.941, < 0.001] (Figure 2). Second, follow-up analyses were carried out on the simple interaction effects to examine the simple main effects. Results showed that for both the elementary and middle school children, normal/word list sentence with natural F0 contours were better recognized than their counterparts with flat F0 contours; normal sentence with natural/flat F0 contours were better recognized than their word list counterparts [elementary school children: F(1,47) > 63.096, p < 0.001, ≥ 0.573; middle school children: F(1,45) > 32.801, p < 0.001, ≥ 0.422].
FIGURE 2
Discussion
The present developmental study explored Mandarin speech recognition by children beyond the age of 9. Two distinct patterns emerged in assessing whether and how age group modulate the influences of low-level F0 contours and high-level semantic context on Mandarin speech recognition in quiet and interference backgrounds. Specifically, there was a significant modulation effect of age group on F0 contours, indicating that the elementary school children relied more on natural F0 contours than the middle school children during Mandarin speech recognition. By contrast, there was no significant modulation effect of age group on semantic context, indicating that children of both age groups used semantic context to assist speech recognition to a similar extent. Furthermore, the significant modulation effect of age group on the interaction between F0 contours and semantic context revealed that younger children could not make better use of semantic context in recognizing speech with flat F0 contours compared with natural F0 contours, while older children benefited more from semantic context when natural F0 contours were altered, thus confirming the important role of F0 contours in Mandarin speech recognition by the elementary school children.
Children’s ability to recognize speech in suboptimal listening backgrounds improves with age (; ), and their ability to use semantic context to assist speech recognition also shows a developmental change from childhood to adolescence (; ). For the present study that focused on elementary and middle school-aged children, our results showed that both groups of children performed better on speech recognition in quiet than in the interference background and that both groups benefited from semantic context to a similar degree regardless of listening backgrounds. These results indicate that Mandarin-speaking children 9.6–10.1 of age (Grade 4) have developed the ability to recognize speech against an interfering talker as well as the ability to use semantic context to assist speech recognition, both of which are comparable with those of children 13.5–14.2 of age (Grade 8). Because only two age groups of children were included in this study, it is unclear at what age children start to develop such abilities comparable with the middle school-aged children.
Although recognition of degraded speech by children has been investigated in several previous studies (e.g., ), how alterations in natural F0 contours affect speech intelligibility for children has rarely been investigated. Natural F0 contours play an important role in speech recognition because dynamic changes in natural F0 contours help listeners recognize voice/unvoiced phonetic segments, separate words in continuous speech and focus attention on the content words. Infants are sensitive to F0 patterns of speech. For example, newborns perceive emotion conveyed via F0 contours in maternal speech () and distinguish between languages based on prosodic information (; ). Even newborns’ cry melody is influenced by F0 patterns of their native languages (). Our finding that recognition of speech with flat F0 contours degraded to a greater extent in the elementary school children than in the middle school children indicate that younger children rely more on natural F0 contours to assist speech recognition. One plausible explanation is that flattening natural F0 contours alters the acoustic-phonetic features of some speech segments and contrasts between words and makes it more difficult for younger children to parse continuous speech into meaningful units, while older children are likely to have developed highly effective and efficient strategies for compensating for the detrimental effects by focusing their attention on the cues that are less vulnerable to F0 distortion. An alternative explanation is that the monotonous speech with flat F0 contours are not attractive enough to the younger listeners, thus resulting in additional costs related to attentional abatement when compared with the older children.
Previous work on young adults shows that F0 contours affect speech recognition interactively with listening background and semantic context. In particular, flattening or inverting F0 contours lowers speech intelligibility to a larger degree in adverse listening backgrounds than in quiet no matter whether the target language is a non-tonal or tonal language; the detrimental effect, however, is more serious for a tonal language (; ; ). In addition, alteration in F0 contours deteriorates speech intelligibility of low-predictability/word list sentences more than high-predictability/normal sentences (; ). In the present study, we found a significant modulation effect of age on the interaction between F0 contours and semantic context, but the interaction between F0 contours and listening background was not modulated by age. Specifically, for the elementary school children, semantic context did not provide extra benefit when F0 contours were flattened compared with speech with natural F0 contours. However, for the middle school children, semantic context did provide extra benefit when F0 contours were flattened. These results seem to indicate that the middle school-aged children could use semantic context to compensate for the degraded speech signal caused by flat F0 contours, while this ability is not well developed in the elementary school children. By contrast, both age groups of children benefited to a similar extent from semantic context when the listening background changes from quiet to interference condition, which indicates that the elementary and middle school children are capable of using semantic context to resist interfering speech. Speech recognition by young and older adults in quiet and single-talker interference backgrounds has been explored in our previous studies ( for young adults, and for older listeners). Although it is unjustifiable to make a direct statistical comparison across the results of different age groups because different word materials and elicitation approaches were adopted, it is meaningful to compare the different patterns of interaction effects between semantic context and F0 contours/listening background across the four age groups. For the middle school children and young adults, interaction effects between semantic context and F0 contours/listening background are both significant, indicating that the two groups could use semantic context to compensate for the distorted speech signal caused by flat F0 contours and interfering speech. Interestingly, for the elementary school children, only the interaction effect between semantic context and listening background is significant; while for the older listeners, only interaction effect between semantic context and F0 contours is significant. The modulation effects of age on the interactions between semantic context and F0 contours/listening background are likely to be related to the maturational/decaying process of speech recognition with degraded signal. Our data indicate that the different age groups may have different receptive skills in processing distortions caused by competing speech and flattened F0 contours. Competing speech results in acoustic and informational masking to the recognition task, while altering F0 contours distorts the target speech signal. That is, competing speech and flattened F0 contours introduces external and internal distortions, respectively, which may need to recruit different cognitive and neural processes during speech perception depending on the listeners’ age. Specifically, it has been suggested that the downstream brain areas (e.g., bilateral inferior frontal regions) are involved in the processing of external distortions, whereas the upstream areas (e.g., left temporal regions) are responsible for the processing of internal distortions (). The age-related differences in the use of semantic context to assist the recognition of different kind of distorted speech might be associated with the developmental/aging changes in cortical/subcortical networks that process the target speech signal in the presence of the two kinds of signal degradation, which calls for further investigation in future studies.
The masker stimuli used in the current study are consonant-misplaced sentences that are syntactically anomalous and semantically incomprehensible, and thus primarily pose energetic (as opposed to informational) masking on the targets. One concern might arise with the specific masking effects on the current findings. Various kinds of maskers (e.g., broadband noise, one- and multi-talker babble) have different acoustic/phonetic properties and thus differ greatly in masking targets and distracting listeners. Further studies using different kinds of interfering sounds are necessary to clarify how children’s speech recognition is affected by characteristics of maskers.
Conclusion
Our current results provide supporting evidence that children use semantic context to assist speech recognition in quiet and adverse listening backgrounds. Although the abilities to recognize speech in adverse listening backgrounds and to use semantic context to assist speech recognition appear to have well developed in normal children aged 9 and beyond, younger children depend more on natural F0 contours in speech recognition than older children. These developmental changes in the influences of semantic context and F0 contours might reflect the differences in auditory and cognitive resources associated with processing of the two types of information in speech perception.
Statements
Author contributions
Research design: HZ, YL, CQG, LZ, and HS. Data collection: HZ, ML, and YL. Data analysis: HZ, YL, and ML. Manuscript preparation: LZ, YZ, and HS.
Acknowledgments
This research was supported by the Open Research Fund of the State Key Laboratory of Cognitive Neuroscience and Learning (CNLYB1607) to LZ, by the National Key Basic Research Program of China (2014CB846103), Natural Science Foundation of China (81461130018, 31671126, 31611130107) and Beijing Municipal Science and Technology Commission (Z151100003915122) to HS, and in part from the National Institute of Child Health and Human Development (P01HD001994) to Haskins Laboratories.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
References
1
AdankP.DavisM. H.HagoortP. (2012). Neural dissociation in processing noise and accent in spoken language comprehension.Neuropsychologia5077–84. 10.1016/j.neuropsychologia.2011.10.024
2
BenichovJ.CoxL. C.TunP. A.WingfieldA. (2012). Word recognition within a linguistic context: effects of age, hearing acuity, verbal ability, and cognitive function.Ear Hear.33250–256. 10.1097/AUD.0b013e31822f680f
3
BinnsC.CullingJ. F. (2007). The role of fundamental frequency contours in the perception of speech against interfering speech.J. Acoust. Soc. Am.1221765–1776. 10.1121/1.2751394
4
BlandyS. E.LutmanM. E. (2005). Hearing threshold levels and speech recognition in noise in 7-year-olds.Int. J. Audiol.44435–443. 10.1080/14992020500189203
5
ChenF.WongL. L.HuY. (2014). Effects of lexical tone contour on Mandarin sentence intelligibility.J. Speech Lang. Hear. Res.57338–345. 10.1044/1092-4388
6
ColeR. A.PerfettiC. A. (1980). Listening for mispronunciations in a children’s story: the use of context by children and adults.J. Verbal Learn. Verbal Behav.19297–315. 10.1016/S0022-5371(80)90239-X
7
CraigC. H.KimB. W.RhynerP. M.ChirilloT. K. (1993). Effects of word predictability, child development, and aging on time-gated speech recognition performance.J. Speech Hear. Res.36832–841. 10.1044/jshr.3604.832
8
CutlerA.DahanD.van DonselaarW. (1997). Prosody in the comprehension of spoken language: a literature review.Lang. Speech40k141–201. 10.1177/002383099704000203
9
DubnoJ. R.AhlstromJ. B.HorwitzA. R. (2000). Use of context by young and aged adults with normal hearing.J. Acoust. Soc. Am.107538–546. 10.1121/1.428322
10
EisenbergL. S.ShannonR. V.MartinezA. S.WygonskiJ.BoothroydA. (2000). Speech recognition with reduced spectral cues as a function of age.J. Acoust. Soc. Am.1072704–2710. 10.1121/1.428656
11
ElliotL. L. (1979). Performance of children aged 9 to 17 years on a test of speech intelligibility in noise using sentence material with controlled word predictability.J. Acoust. Soc. Am.66651–653. 10.1121/1.383691
12
FallonM.TrehubS. E.SchneiderB. A. (2000). Children’s perception of speech in multitalker babble.J. Acoust. Soc. Am.1083023–3029. 10.1121/1.1323233
13
FernaldA. (2001). Making use of semantic context in early language understanding.Paper Presented at the Society for Research in Child Development, Minneapolis, MN.
14
FernaldA.SimonT. (1984). Expanded intonation contours in mothers’ speech to newborns.Dev. Psychol.20104–113. 10.1037/0012-1649.20.1.104
15
HallJ. W.GroseJ. H.BussE.DevM. B. (2002). Spondee recognition in a two-talker masker and a speech-shaped noise masker in adults and children.Ear Hear.23159–165. 10.1097/00003446-200204000-00008
16
HuaZ.DoddB. (2000). The phonological acquisition of Putonghua (modern standard Chinese).J. Child Lang.273–42.
17
JiangW.LiY.HuaS.ZhangL.ZhangY. (2017). Use of semantic context and F0 contours by older listeners during Mandarin speech recognition in quiet and single-talker interference conditions.J. Acoust. Soc. Am.141EL338–EL344. 10.1121/1.4979565
18
LauresJ. S.BuntonK. (2003). Perceptual effects of a flattened fundamental frequency at the sentence level under different listening conditions.J. Commun. Disord.36449–464. 10.1016/S0021-9924(03)00032-7
19
MampeB.FriedericiA. D.ChristopheA.WermkeK. (2009). Newborns’ cry melody is shaped by their native language.Curr. Biol.191994–1997. 10.1016/j.cub.2009.09.064
20
MaxwellS. E.DelaneyH. D. (2004). Designing Experiments and Analyzing Data: A Model Comparison Perspective, 2nd Edn. Mahwah, NJ: Erlbaum.
21
MillerS. E.SchlauchR. S.WatsonP. J. (2010). The effects of fundamental frequency contour manipulations on speech intelligibility in background noise.J. Acoust. Soc. Am.128435–443. 10.1121/1.3397384
22
MooreB. C. J.TylerL. K.Marslen-WilsonW. (2008). Introduction. The perception of speech: from sound to meaning.Philos. Trans. R. Soc. Lond. B Biol. Sci.363917–921. 10.1098/rstb.2007.2195
23
NazziT.FlocciaC.BertonciniJ. (1998). Discrimination of pitch contours by neonates.Infant. Behav. Dev.21779–784. 10.1016/S0163-6383(98)90044-3
24
PapsoC. F.BloodI. M. (1989). Word recognition skills of children and adults in background noise.Ear Hear.10235–236. 10.1097/00003446-198908000-00004
25
PatelA. D.XuY.WangB. (2010). “The role of F0 variation in the intelligibility of Mandarin sentences,” inProceedings of Speech Prosody 2010, Chicago, IL.
26
Pichora-FullerM. K.SchneiderB. A.DanemanM. (1995). How young and old adults listen to and remember speech in noise.J. Acoust. Soc. Am.97593–608. 10.1121/1.412282
27
RamusF.HauserM. D.MillerC.MorrisD.MehlerJ. (2000). Language discrimination by human newborns and by cotton-top tamarin monkeys.Science288349–351. 10.1126/science.288.5464.349
28
ScottS. K.RosenS.WickhamL.WiseR. J. (2004). A positron emission tomography study of the neural basis of informational and energetic masking effects in speech perception.J. Acoust. Soc. Am.115813–821. 10.1121/1.1639336
29
SheldonS.Pichora-FullerM. K.SchneiderB. A. (2008). Priming and sentence context support listening to noise-vocoded speech by younger and older adults.J. Acoust. Soc. Am.123489–499. 10.1121/1.2783762
30
SmiljanicR.SladenD. (2013). Acoustic and semantic enhancements for children with cochlear implants.J. Speech Lang. Hear. Res.561085–1096. 10.1044/1092-4388(2012/12-0097)
31
StelmachowiczP. G.HooverB. M.LewisD. E.KortekaasR. W. L.PittmanA. L. (2000). The relation between stimulus context, speech audibility, and perception for normal hearing and hearing-impaired children.J. Speech Lang. Hear. Res.43902–914. 10.1044/jslhr.4304.902
32
StuartA. (2005). Development of auditory temporal resolution in school-age children revealed by word recognition in continuous and interrupted background noise.Ear Hear.2678–88. 10.1097/00003446-200502000-00007
33
StuartA.GivensG. D.WalkerL. J.ElangovanS. (2006). Auditory temporal resolution in normal-hearing preschool children revealed by word recognition in continuous and interrupted noise.J. Acoust. Soc. Am.1191946–1949. 10.1121/1.2178700
34
WangJ.ShuH.ZhangL.LiuZ.ZhangY. (2013). The roles of fundamental frequency contours and sentence context in Mandarin Chinese speech intelligibility.J. Acoust. Soc. Am.34EL91–EL97. 10.1121/1.4811159
35
WangW. S. Y. (1973). The Chinese language.Sci. Am.22850–60. 10.1038/scientificamerican0273-50
36
WermkeK.RuanY.FengY.DobnigD.StephanS.WermkeP.et al (2016). Fundamental frequency variation in crying of Mandarin and German neonates.J. Voice31255.e25–255.e30. 10.1016/j.jvoice.2016.06.009
37
WilsonR. H.FarmerN. M.GandhiA.ShelburneE.WeaverJ. (2010). Normative data for the words-in-noise test for 6- to 12-year-old children.J. Speech Lang. Hear. Res.531111–1121. 10.1044/1092-4388(2010/09-0270)
38
XuG.ZhangL.ShuH.WangX.LiP. (2013). Access to lexical meaning in pitch-flattened Chinese sentences: an fMRI study.Neuropsychologia51550–556. 10.1016/j.neuropsychologia.2012.12.006
39
ZhangL.LiY.WuH.LiX.ShuH.ZhangY.et al (2016). Effects of semantic context and fundamental frequency contours on Mandarin speech recognition by second language learners.Front. Psychol.7:908. 10.3389/fpsyg.2016.00908
Summary
Keywords
semantic context, fundamental frequency contours, speech recognition, interfering speech, children
Citation
Zhou H, Li Y, Liang M, Guan CQ, Zhang L, Shu H and Zhang Y (2017) Mandarin-Speaking Children’s Speech Recognition: Developmental Changes in the Influences of Semantic Context and F0 Contours. Front. Psychol. 8:1090. doi: 10.3389/fpsyg.2017.01090
Received
12 March 2017
Accepted
13 June 2017
Published
28 June 2017
Volume
8 - 2017
Edited by
Chia-Ying Lee, Academia Sinica, Taiwan
Reviewed by
Gang Peng, The Chinese University of Hong Kong, Hong Kong; Jing Zhao, Capital Normal University, China
Updates
Copyright
© 2017 Zhou, Li, Liang, Guan, Zhang, Shu and Zhang.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Linjun Zhang, zhanglinjun75@gmail.com
This article was submitted to Language Sciences, a section of the journal Frontiers in Psychology
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.