The present study investigates the so-called frequency inheritance effect in word production. According to some earlier studies (e.g. Jescheniak & Levelt, 1994), retrieval of a low-frequency homophone benefits from its high-frequency homophone twin, and more specifically word-retrieval RT is determined by the frequency of the phonological form of the word (sum of homophone frequencies) rather than the frequency of the specific word. This result, however, has been challenged by later studies (e.g. Caramazza et al., 2001) and one possible resolution is that languages differ in the extent to which the inheritance effect occurs. Two experiments are reported to test whether the frequency inheritance effect depends on the target language, namely, if a language such as Korean with relatively many homophones tend not to show frequency inheritance, which is compared with the language with fewer homophones such as Dutch and German (Jescheniak & Levelt, 1994; Jescheniak et al., 2003). Experiment 1 was picture naming, and Experiment 2 used an English-to-Korean translation task. In both experiments, the homophones were actually slower than the low-frequency controls, suggesting that there was no evidence for the inheritance effect. These results imply that the issue of whether specific word or homophone frequency determines production can be properly assessed by taking into account the language-specific nature of the lexicon such as the percentage of the homophone words in that language.

This paper attempts to investigate phonetic substances of the stress patterns of English compounds and noun phrases, showing that the theoretically derived stress structures are not consistent with the accentual patterns in real utterances. Even though it has been long claimed that compounds have the stress pattern [1 3] and that noun phrases, [2 1] as in Chomsky & Halle (1968), their difference has not been yet explored empirically or phonetically. I present a phonetic experiment conducted to see if there is any difference along the tonal contours, mostly focusing on their pitch accent distribution. 36 different compounds and 36 different noun phrases included in carrier sentences were examined, and they were varied in position within a sentence. Results showed that various accentual patterns were produced, and among them, [H* X] predominantly occurs in all three positions in both compounds and noun phrases, whereas the patterns [X H*] and [X X] appear relatively more frequently in final position than in initial and medial position. Furthermore, the pattern [Ac + No], in which the preceding element is pitch-accented with no accent on the following one, is the major stress pattern in both compounds and noun phrases and in all three sentence positions. This suggests that there seems to be no difference in accentual patterns between compounds and noun phrases, which is not consistent with the hypothesis. The results are interpreted as saying that the preceding element alone tends to be prominent with no accent following it both in compounds and noun phrases, and that therefore, theoretically speculated phonological claims are not always phonetically supported.

The current study examines the role of post-lexical tonal patterns of a prosodic phrase in word segmentation. In a word spotting experiment, native Korean listeners were asked to spot a disyllabic or trisyllabic word from twelve syllable speech stream that was composed of three Accentual Phrases (AP). Words occurred with various post-lexical intonation patterns. The results showed that listeners spotted more words in phrase-initial than in phrase-medial position, suggesting that the AP-final H tone from the preceding AP helped listeners to segment the phrase-initial word in the target AP. Results also showed that listeners' error rates were significantly lower when words occurred with initial rising tonal pattern, which is the most frequent intonational pattern imposed upon multisyllabic words in Korean, than with non-rising patterns. This result was observed both in AP-initial and in AP-medial positions, regardless of the frequency and legality of overall AP tonal patterns. Tonal cues other than initial rising tone did not positively influence the error rate. These results not only indicate that rising tone in AP-initial and AP_final position is a reliable cue for word boundary detection for Korean listeners, but further suggest that phrasal intonation contours serve as a possible word boundary cue in languages without lexical prominence.

In this study, the researchers evaluate the hypothesis that the reaction time (RT) in the brain is largely dependent on age and gender. The researchers tested English and Korean words using two types: reading as visual stimuli, and listening as auditory stimuli. The E-Prime program installed in functional magnetic resonance imaging (f-MRI) was used. There are 136 Korean subjects in three groups: 30 children (5th and 6th grade elementary school students), 76 young adults (college students), and 30 adults (35-53 years old). In the listening test, the variation for the two languages was different among the three age groups, regardless of gender, whereas the data did not show meaningful differences in the reading test. The findings will provide some meaningful information regarding perception and acquisition of a foreign language.

This study explored the perception of transplanted English prosody by thirty American and Korean, male and female listeners. The English utterances of various sentence types produced by Korean and American male speakers were employed to transplant the American prosody contours to Korean English utterances. Then, the thirty subjects were instructed to rate the transplanted prosodic components. Results showed that the interactions between the three factors (e.g., rater groups & transplantation types; transplantation types & sentence types; rater groups & transplantation types & sentence types) turned out to be meaningful. Both Americans and Koreans perceived the effectiveness of the combined effect of transplanted duration and pitch or duration and pitch and intensity. However, when perceiving individual prosodic components, Americans and Koreans showed different perceptual ratings. As for the overall prosody change, Americans perceived the change of intensity in a significant way but Koreans did not because intensity is not a crucial semantic factor in Korean. Americans rated the transplantation of duration alone as ineffective while Koreans rated otherwise. This was explained by the difference between English and Korean. The difference of perspective was also significant with different sentence types, especially with the three sentence types that had speech rates slower than other sentence types. A slower speech rate intensified the mismatch between the transplanted duration and the original pitch causing a negative impression on American listeners whereas this did not affect Korean listeners. Pedagogical implications of the findings are discussed.

The aims of this study were to examine the acoustic features of Korean and Japanese vowels, and to determine whether new phones that do not have counterparts in Japanese or similar phones that have counterparts improve more from learning. This study consisted of three parts. In Experiment I, a speech production test was performed to observe the acoustic features of Korean and Japanese vowels. In Experiment II, the speech production of Korean vowels produced by Koreans, advanced Japanese learners of Korean, and beginning Japanese learners of Korean was investigated. In Experiment III, a speech perception study of Korean vowels produced by the two Japanese learner groups was conducted to observe the effect of learning on acquiring L2 phones. The conclusion drawn from the study was that the similar phones produced by Japanese show more similarity with those of Koreans than new phones in terms of F1 and F2, but Japanese learners of Korean displayed more improvement in new phones from learning.

The purpose of this study is to examine Korean students' production and perception of word-final stop voicing in light of their overseas experience. Subjects were English native speakers, Korean university students with residence experience in America, Korean university students without residence experience in America, and Korean elementary school students. They participated in both production and perception tests. Results showed that the students' production and perception with residence experience in America appeared quite similar to those of the English native speakers. In the production tests, we noticed somewhat different results in temporal and frequency features. The one-year residence in America had some influence on their frequency features, but not the temporal features in the word final stop production. That difference could be seen in the perception tests, too. We could not find any difference in the identification test of the final release environment between the Korean university students who had studied abroad and those who didn't. Rather the difference could be found in the cue influence test in both the final release and non-release environments.

The purpose of this paper is to examine how Korean speakers realize English stress and rhythm at the sentence level, and investigate what different acoustic characteristics of English sentence stress and rhythm Korean speakers have, compared with those of American English speakers. Stressed words in the sentence were analyzed in terms of duration, fundamental frequency, and intensity of the stressed vowel in the word with neutral stress and with emphatic stress, respectively. According to the results, when the words had emphatic stress, both Koreans' and Americans' F0 and intensity of the stressed vowel were higher than those with neutral stress. Korean speakers of English realized the sentence stress with shorter vowel duration and higher F0 than American English speakers when the words had emphatic stress. The analysis of the timing of the sentence with increased unstressed syllables showed that both Americans and Koreans produced the sentence with longer duration as the number of unstressed syllables increased. However, the duration of unstressed syllables between stressed syllables by Koreans was longer than that by Americans. Americans seemed to produce unstressed syllables between stressed syllables faster than Koreans for regular intervals of stressed syllables. This analysis implies that if there are more unstressed syllables between stressed syllables, Koreans might produce unstressed syllables and the whole sentence with longer duration.

The purpose of this study was to examine the syllable structures of ten Korean numeric sounds produced by ten students. Each sound was normalized by its maximum intensity value and divided into onset, vowel, and coda sections after finding abrupt or visible changes in energy values or cumulative values of lower spectral energy at each pulse point using four Praat scripts. Then, segmental durations and cumulative intensity values of each syllable were obtained to find a statistical summary of the syllable structure. Intensity values at 100 proportional time points were also collected to compare the ten sounds. Results showed as follows: Firstly, there was not much deviation from the grand average duration and intensity for the majority of the sounds except the two diphthongal sounds on which their boundary points varied among the speakers. Secondly, the onset point for the CV or CVC category sounds and the boundary between the vowel and the nasal or lateral sound were easy to identify, which may be automatically traced later. Thirdly, there seems some tradeoff among the sections maintaining the same total duration per each syllable. Further studies on syllables with various onsets or codas would be desirable to make a general statement on the Korean syllable structure.

This study focuses on the classification of pathological voice using GMM (Gaussian Mixture Model) and compares the results to the previous work which was done by ANN (Artificial Neural Network). Speech data from normal people and patients were collected, then diagnosed and classified into two different categories. Six characteristic parameters (Jitter, Shimmer, NHR, SPI, APQ and RAP) were chosen. Then the classification method based on the artificial neural network and Gaussian mixture method was employed to discriminate the data into normal and pathological speech. The GMM method attained 98.4% average correct classification rate with training data and 95.2% average correct classification rate with test data. The different mixture number (3 to 15) of GMM was used in order to obtain an optimal condition for classification. We also compared the average classification rate based on GMM, ANN and HMM. The proper number of mixtures on Gaussian model needs to be investigated in our future work.

All observation has uncertainty due to noise or channel characteristics. This uncertainty should be counted in the modeling of observation. In this paper we propose a modified optimization object function of a GMM training considering inexact observation. The object function is modified by introducing the concept of observation confidence as a weighting factor of probabilities. The optimization of the proposed criterion is solved using a common EM algorithm. To verify the proposed method we apply it to the speaker recognition domain. The experimental results of text-independent speaker identification with VidTimit DB show that the error rate is reduced from 14.8% to 11.7% by the modified GMM training.