Shigeaki Amano

Infant-directed speech (IDS), which is spoken to an infant by a caretaker such as a parent, was investigated in terms of its longitudinal change in a speaking rate using the NTT infant speech database. It was revealed that IDS speaking rate was slow and it gradually became faster when infant's was 0 to about 25-months old, but it was almost constant after about 25-months old. Considering that I...

NTT Infant Database was developed from longitudinal recordings of utterances of infants and their parents. It is a huge database with many kinds of information such as an utterance transcription, a phoneme label, and a fundamental frequency. We describe the developmental procedure and contents of this database.

Familiarity-controlled Word-lists 2007 (FW07) has been released. FW07 was developed for assessing hearing ability in daily life conditions. FW07 aims to be easier use in clinical examinations, such as fitting hearing aid. For this aim, we made two modifications to FW03. First, the number of word in each list was reduced from 50 to 20 to lessen subjects' burden. Second, not only the phonemic dis...

Familiarity-controlled Word-lists 2007 (FW07) has been released. FW07 was developed for assessing hearing ability in daily life conditions. FW07 aims to be easier use in clinical examinations, such as fitting hearing aid. For this aim, we made two modifications to FW03. First, the number of word in each list was reduced from 50 to 20 to lessen subjects' burden. Second, not only the phonemic dis...

Familiarity-controlled Word-lists 2007 (FW07) has been released. FW07 was developed for assessing hearing ability in daily life conditions. FW07 aims to be easier use in clinical examinations, such as fitting hearing aid. For this aim, we made two modifications to FW03. First, the number of word in each list was reduced from 50 to 20 to lessen subjects' burden. Second, not only the phonemic dis...

Japanese case particle ga indicate the following three relationships between noun phrase and predicate: NOMINATIVE-MOVEMENT (inu ga hoeru [dog NOM bark; 'a dog barks']); NOMINATIVE-CONDITION (inu ga iru [dog NOM exist; 'there is a dog']); and ACCUSATIVE-CONDITION (inu ga hoshii [dog ACC want; 'I want a dog']). To examine the order of occurrence of the three relationships, the analyses were perf...

This paper shows the degrees to which hearing ability and language ability influence the spoken word intelligibility of aged people. Thirty-two subjects (14 males and 18 females) participated in a spoken word intelligibility test with an adiometory test, the vocabulary test in Wechsler Adult Intelligence Scale-Third Edition (WAIS-III), and 100RAKAN (Reading Ability test for KANji words). The su...

This paper shows the degrees to which hearing ability and language ability influence the spoken word intelligibility of aged people. Thirty-two subjects (14 males and 18 females) participated in a spoken word intelligibility test with an adiometory test, the vocabulary test in Wechsler Adult Intelligence Scale-Third Edition (WAIS-III), and 100RAKAN (Reading Ability test for KANji words). The su...

This paper shows the degrees to which hearing ability and language ability influence the spoken word intelligibility of aged people. Thirty-two subjects (14 males and 18 females) participated in a spoken word intelligibility test with an adiometory test, the vocabulary test in Wechsler Adult Intelligence Scale-Third Edition (WAIS-III), and 100RAKAN (Reading Ability test for KANji words). The su...

A previous study (Mugitani et al. 2004) indicated that 8-month-old Japanese infants match the phonetic information of vowel /a/ to its respective visible lip articulation, while such audiovisual matching of lips and voice was not evidenced for vowel /i/. This study aimed at further verifying the development of audiovisual lip-voice matching of vowel HI in Japanese infants. Experiment 1 indicate...

In this paper, we quantitatively investigate the mechanism of children's deviations from adult articulation forms(CDFA) by using an articulation control model. The target CDFA is the substitution of /sh/ for /s/ and /n/ for /d/ at the beginning of words or after prolonged sounds, and the substitution of /r/ for /d/ in the middle of words. Our analysis shows that the CDFA at the beginning of wor...

In this paper we present the motivation for the construction of the Hinoki treebank. It is a rich and dynamic treebank of dictionary definition sentences parsed using a Japanese HPSG. We show how the treebank is being used to build an ontology, and outline plans for further work.

In this paper we describe the construction of the Japanese Semantic Lexicon: Lexeed. The lexicon contains the most familiar 28,000 thousand Japanese words, which also form the defining vocabulary for their definitions. The lexicon is designed to be a base for further information processing.

A set of 44 semantic and 4 phonological word-fluency tasks were administered to 50 young and 50 healthy older adults. The participant was required to orally produce words in a semantic category, or words beginning with a letter (representing Japanese mora) displayed on a computer screen. She/he was encouraged to utter as many words as possible in 30 seconds. The result showed that the performan...

A set of 44 semantic and 4 phonological word-fluency tasks were administered to 50 young and 50 healthy older adults. The participant was required to orally produce words in a semantic category, or words beginning with a letter (representing Japanese mora) displayed on a computer screen. She/he was encouraged to utter as many words as possible in 30 seconds. The result showed that the performan...

A set of 44 semantic and 4 phonological word-fluency tasks were administered to 50 young and 50 healthy older adults. The participant was required to orally produce words in a semantic category, or words beginning with a letter (representing Japanese mora) displayed on a computer screen. She/he was encouraged to utter as many words as possible in 30 seconds. The result showed that the performan...

As the first step of constructing a dictionary of word concepts, the "Commonsense Concept Database," which will be a base for language processing technologies regarding meaning, we selected basic words which are supposed to be commonly used by Japanese adults. We selected the basic words from a Japanese dictionary in which the number of word entries is about 95,000. In a previous study, the siz...

As the first step of constructing a dictionary of word concepts, the "Commonsense Concept Database," which will be a base for language processing technologies regarding meaning, we selected basic words which are supposed to be commonly used by Japanese adults. We selected the basic words from a Japanese dictionary in which the number of word entries is about 95,000. In a previous study, the siz...

Comprehension and production of spoken and written words are affected by various word attributes such as word frequcy, familiarity, and imageability(subjective ratings of the ease eith which the referent of a word evokes its mental image). This study intends to rate imageability for a large number of aurally and visually presented Japanese words. The correlation coefficient between visual and a...

Databases of lexical properties of Japanese words and characters were developed. The database, Nihongo-no Goitokusei(Lexical Properties of Japanese), consists of seven volumes. Volumes 1 through 6 [ref's 1-6]include subjective rating data for word familiarity and orthographic plausibility for 80, 000 words in the Shinmeikai Japanese Dictionary[ref 7], as well as character familiarity ratings an...

NTT psycholinguistic databases "Lexical Properties of Japanese" were developed for the large number of Japanese words and characters. The databases contain word and character information such as familiarity, frequency of occurrence, appropriateness of accent, appropriateness of orthography, subjective complexity, and other important psycholinguistic properties. Words and characters satisfying m...

Previous studies of phonetic cueing of focus structure have indicated that the role of fundamental frequency(F0). The current study examined the role of F0 in focus perception. Subjects select the word emphasized by a speaker for each stimulus, in which the F0 values were changed at three points:the F0 minimum point (dip) between the two words in a noun phrase, and the F0 maximum point (peak) i...

Functional load of pitch accent is proposed for Japanese words. It is calculated from 1) the ratio of the number of a pair of homophones with opposing pitch accent to the number of those with non-opposing pitch accent, and 2) word familiarity of the homophones. The functional load is larger when the difference of word familiarity scores is smaller between homophones with opposing pitch accent. ...

The perceptual effects of lexical stratum (Sino-Japanese or Foreign etymology) were investigated. Japanese listeners judged the [a]-[aa] boundary using PEST. The vowels were in carrier nonwords of the form [CoC^*a(a)]. The consonants C and C' were varied to make 9 carriers, from the Foreign-sounding [pofa(a)] to the Sino-Japanese-sounding [ryohya(a)]. Foreign words often end in [aa], while Sino...

Distributions of word length and mora of 41,789 Japanese spoken words were analyzed according to auditory word familiarity. It was revealed that relative frequencies of word length and mora have significant differences among high-, medium-, and low-word-familiarity group. For example, relative frequencies of 2-mora-long words, /ru/, /su/, and special mora of /Q/ in high-word-familiarity group a...

Brain activity was measured by the whole head type SQUID when subjects made a lexical decision totime-compressed, normal, or time-expanded words and nonwords. Single dipole model was used to estimate the source location of the brain megnetic field which corresponds to the word perceptual process. It was located to the auditory area in the left hemisphere.

To investigate the time-course of phoneme perception and word perception, an occurrence point of a phoneme, a phoneme detection point, and a lexical decision point were measured for time-modified spoken Japanese. It was revealed that the phoneme detection point is predictable by a linear function of the occurrence point of the phoneme regardless of the time-modified rate, and that a constant nu...

Reaction times and rate of phonemic restoration of a stop consonant within VCV syllables were measured to investigate perceptual integration of distributed information of the stop consonant in pre- and post closure parts by speeded and non- speeded perceptual experiments.It was revealed that perceptual processes both in the speeded and non-speeded experiments used the distributed information in...