This study explicitly tested listeners' categorization of stops preceding vowels with varying durations, which will help establish how Korean listeners' contrastive perception operates along durational cues.īeyond duration, we also examine the possibility of the onset F0 cue being relativized to the offset F0 to inform stop categorization. It is conceivable that the perception of stop closure in Korean arises from a temporal relativization with locally adjacent segments. Shortening of voice onset time in Korean aspirates: Phonetic OCP or timing regulation?,” in Proceedings of the 19th International Congress of Phonetic Sciences, August 5–9, Melbourne, Australia, pp. Ithaca, NY.], but vowel duration and stop closure tend to covary inversely: fortis stops are preceded by vowels with short duration, whereas lenis are preceded by vowels with long duration ( Oh and Kim, 2019 13. The phonetics and phonology of stop lenition in Korean,” Ph.D. Not only are there large differences in absolute closure duration between fortis and lenis stops [e.g., Han (1996) 6. First, stop closure may be perceived in relation to the neighboring vowel duration. Ĭapitalizing on the phonetic properties of the two stops, the present study investigates whether stop categorization is driven by a contrastive perception of those cues. Taken together, the fortis and lenis stops can be represented by the following cue pairings, respectively: and. Cho, T., andĪrticulatory and acoustic studies on domain-initial strengthening in Korean,” J. However, the two stops can also be characterized by F0: regardless of position, vowels following fortis stops have a high F0 at the onset, whereas the onset of those coming after lenis stops has a low F0 ( Cho and Keating, 2001 3. The phonetics and phonology of ‘tense’ and ‘plain’ consonants in Korean,” Ph.D. The two stops are primarily distinguished by closure duration, with fortis stops having considerably longer closure than lenis stops [ M(fortis) = 140 ms vs M(lenis) = 56 ms Han (1996) 6. This inquiry was tested against how phonetic cues to the fortis-lenis distinction are processed by Korean listeners in a word-medial position. In particular, learners of a tonal second language may process a perceptual contrast along F0 cues for non-native stop identification. The results highlight the importance of language-specific linguistic knowledge of the sounds: lip rounding ordinarily involved in both /u/ and /y/ in French may have caused the two sounds to be grouped into a similar category, eliciting a robust perceptual contrast with neighboring fricatives.īuilding upon this line of research, the present study explores how experience in a second language may build perceptual contrasts along newly acquired phonetic cues.
![onset stop closure durations onset stop closure durations](https://www.e-ceo.org/upload/thumbnails/ceo-2016-00626f1.gif)
However, such a perceptual contrast for the /y/ vowel, despite its relatively high F2, was only observed for French-speaking listeners, while English-speaking listeners treated /y/ and /a/ similarly. The results indicate that the low F2 of the neighboring caused the listeners to perceive the spectrally higher fricative. , for example, reported that both English and French native speakers were more likely to categorize ambiguous noise signals as /s/ before /u/ than before /a/. Kang, S.,Įffects of native language on compensation for coarticulation,” Speech Commun. ).Ĭontrastive perception is further modulated by linguistic experience. Relational timing in the production and perception of Japanese singleton and geminate stops,” Phonetica 67, 25– 46. Similarly, Japanese listeners are biased toward perceiving “geminates,” or long consonants, when the preceding mora has been shortened (e.g., Idemaru and Guion-Anderson, 2010 7. Preceding vowel duration as a cue to the perception of the voicing characteristics of word-final consonants in English,” J. Vowel-length differences before voiced and voiceless consonants: An auditory explanation,” J. English listeners, for instance, perceive a stop segment as “voiceless,” which is characterized by a long stop closure, when the preceding vowel has been shortened [e.g., Kluender et al. In a similar vein, the duration of stop closure is assessed in relation to the duration of a neighboring vowel. L.,Įxtrinsic context affects perceptual normalization of lexical tone,” J. Cantonese speakers were reported to identify an ambiguous level tone as high (/ji 55/ “doctor”) in a context where the F0 of the sentence was lowered ( low-F0 context), while the physically identical tone was identified as low (/ji 22/ “two”) in a context where the F0 of the sentence was raised ( high-F0 context) ( Francis et al., 2006 5. The effect of extrinsic context is said to be “contrastive.” For example, the processing of lexical tone is context-dependent for speakers of tone languages.
![onset stop closure durations onset stop closure durations](https://image.slideserve.com/1109997/english-stop-contrasts1-l.jpg)
The perception of speech sounds arises from the integration of the intrinsic properties of target sounds and the extrinsic properties of neighboring sounds.