Explore the vast range of titles available.

Numeral Classifier Systems considers the functional significance of the Japanese numeral system, its conclusions based on a corpus of 500 uses of classifier constructions drawn from oral and written Japanese texts.Interestingly, although the Japanese system appears to conform at least superficially to universalistic predictions about its semantic structure, this study reports that in actual usage, the semantic role of classifiers is slight - only very rarely do they carry any lexical information unavailable from the context or the noun with which the classifier occurs. It does appear, however, that the system has an important role to play in providing pronoun-like anaphoric elements and in marking pragmatic distinctions such as the individuatedness of referents and the newness of numerical information. For these reasons, the classifier system is deeply involved in a number of subsystems of Japanese grammar, and the demise of the system (sometimes rumored to be impending) would have substantial implications for the structure of the language as a whole.

The book is about the numeral classifier system and the acquisition of Japanese classifiers by Japanese children. It consists of two parts. First, it provides a general typological characterization of numeral classifier phrases and discusses problems in determining what constitutes the nature of classifiers. It also discusses the semantic properties of numeral classifiers based on an analysis of four languages from four different language families. Second, it examines the acquisitions of Japanese numeral classifiers by Japanese preschool children, ages 3 to 6, with a primary emphasis on the development of comprehension. The importance of the study is that it reveals that young children have a much greater sensitivity to the conceptual underpinnings of the numeral classifier system than was previously considered to be the case. The research results also provide a converging source of evidence that young children often come to initially grasp the structure of the world in ways that are better understood in cognitive than perceptual terms. The implications will contribute to not only the area of language acquisition but also categorization and conceptual development.

In the first half of 2023, text-generative artificial intelligence (AI), including ChatGPT from OpenAI, has attracted considerable attention worldwide. In this study, first, we compared Japanese stylometric features of texts generated by ChatGPT, equipped with GPT-3.5 and GPT-4, and those written by humans. In this work, we performed multi-dimensional scaling (MDS) to confirm the distributions of 216 texts of three classes (72 academic papers written by 36 single authors, 72 texts generated by GPT-3.5, and 72 texts generated by GPT-4 on the basis of the titles of the aforementioned papers) focusing on the following stylometric features: (1) bigrams of parts-of-speech, (2) bigram of postpositional particle words, (3) positioning of commas, and (4) rate of function words. MDS revealed distinct distributions at each stylometric feature of GPT (3.5 and 4) and human. Although GPT-4 is more powerful than GPT-3.5 because it has more parameters, both GPT (3.5 and 4) distributions are overlapping. These results indicate that although the number of parameters may increase in the future, GPT-generated texts may not be close to that written by humans in terms of stylometric features. Second, we verified the classification performance of random forest (RF) classifier for two classes (GPT and human) focusing on Japanese stylometric features. This study revealed the high performance of RF in each stylometric feature: The RF classifier focusing on the rate of function words achieved 98.1% accuracy. Furthermore the RF classifier focusing on all stylometric features reached 100% in terms of all performance indexes (accuracy, recall, precision, and F1 score). This study concluded that at this stage we human discriminate ChatGPT from human limited to Japanese language.

This study examined the acquisition of numeral classifiers in 120 monolingual Japanese children. Previous research has argued that the complex semantic system underlying classifiers is late acquired. Thus, we set out to determine the age at which Japanese children are able to extend the semantic properties of classifiers to novel items/situations. Participants completed a comprehension task with a mouse-tracking extension and a production task with nonce and familiar items. While the comprehension results showed ceiling effects on familiar and nonce items, age significantly modulated a difference in accuracy between familiar and nonce items in the production task. The findings suggest that the acquisition of the underlying semantic system is acquired much earlier than previously argued. Previously attested issues with Japanese classifier production in young(er) children are more likely to reflect accessing difficulties than indexing the underlying grammatical competence of the classifier system.

One of the central issues in cognition is identifying universal and culturally specific patterns of thought. In this study, we examined how one aspect of culture, a linguistic part of speech known asclassifiers, are related to categorization of solid objects. In Experiment 1, we used a numeral classifier elicitation task to examine the classifiers used by speakers of Hmong, Japanese, and Mandarin Chinese (N = 34) with 135 nouns that referred to solid objects. In Experiment 2, adult speakers of English, Japanese, Mandarin Chinese, and Hmong (N = 64) rated the similarity of 39 pictured objects that depicted a subset of the nouns. All groups classified the objects into natural kinds and artifacts, with the category of humans anchoring both divisions. The main difference that emerged from the study was that speakers of Japanese and English rated humans and animals as more similar to each other than Hmong speakers; Mandarin speakers’ ratings of the similarity between humans and animals fell in between those of Hmong and English speakers. However, the pattern of categorization of humans and animals found among speakers of the classifier languages contradicted their patterns of classifier use. The findings help to tease apart the effects of language from other cultural factors that impact cognition.

The current study examined the comprehension and production of classifiers, case marking, and morphological passive structures among 414 child Japanese heritage speakers (mean age = 10.01 years; range = 4.02 – 18.18). Focusing on individual differences, we extracted latent experiential factors via the Q-BEx questionnaire (De Cat, Kašćelan, Prévost, Serratrice, Tuller, Unsworth, & The Q.-Be Consortium, 2022), which were then used to predict knowledge and use of these grammatical structures. The findings reveal that: (i) experiential factors such as heritage language (HL) engagement at home and within the community modulate grammatical performance differentially from childhood through adolescence, and (ii) HL proficiency, immersion experiences, and literacy systematically predict HL grammatical outcomes. These results indicate that particular language background factors hold differential significance at distinct developmental stages and that higher proficiency, richer immersion experiences, and literacy engagement in the HL are crucial for the development of core grammatical structures.

Models of nominal structure in Japanese and Korean (JK) are commonly built on the assumption that the nominal domain must be head-final because JK clauses show head-final ordering, rather than being directly supported by observable empirical head-final patterns. In order to produce the surface orders that are found in JK nominals, all head-final analyses require massive hidden movements from underlying structures which are never overtly realized in any surface sequencing. This paper suggests that a much more parsimonious analysis of JK nominal structure is available if JK are not taken to be uniformly head-final in their syntax but exhibit a degree of mixed-headedness, as found in German, Hindi, Hixkaryana, Amharic, Persian and other languages. The paper develops such an analysis, in which underlying head-initial structures do occur in surface syntax, and refines this further with support from the patterning of various numeral-classifier-noun relations in Japanese and Korean. The resulting analysis proposes that the functional structure of JK nominals is head-initial, while the lexical domain (nP, NP) is head-final. Such mixed-headedness is shown to accord with the Final Over Final Constraint/FOFC, and hence is not an unconstrained departure from the pure head-finality widely assumed for JK.

With the growing demand for natural and intelligent human–machine interaction in multilingual environments, automatic language identification (LID) has emerged as a crucial component in voice-enabled systems, particularly in the automotive domain. This study proposes an audio-based LID model that identifies the spoken language directly from voice input without requiring manual language selection. The model architecture leverages two types of feature extraction pipelines: a Variational Autoencoder (VAE) and a pre-trained Wav2Vec model, both used to obtain latent speech representations. These embeddings are then fed into a multi-layer perceptron (MLP)-based classifier to determine the speaker’s language among five target languages: Korean, Japanese, Chinese, Spanish, and French. The model is trained and evaluated using a dataset preprocessed into Mel-Frequency Cepstral Coefficients (MFCCs) and raw waveform inputs. Experimental results demonstrate the effectiveness of the proposed approach in achieving accurate and real-time language detection, with potential applications in in-vehicle systems, speech translation platforms, and multilingual voice assistants. By eliminating the need for predefined language settings, this work contributes to more seamless and user-friendly multilingual voice interaction systems.

While many languages require an obligatory plural morpheme to make reference to plural individuals, numeral classifier languages generally do not (Greenberg 1972; Sanches & Slobin 1973; Doetjes 2012: a.o.). This led some researchers to conclude that noun denotations in numeral classifier languages are inherently plural. In this paper, I show that one can find a syntactic environment which requires a plural reading or a mass reading of a noun phrase in a numeral classifier language. The core empirical finding is that the postnominal measurement construction in Japanese (i) does not allow a singular reading, and (ii) sometimes triggers count-to-mass coercion. This suggests that the constraint that measure phrases select a non-quantised denotation (Krifka 1989) is non-trivially satisfied in Japanese even though Japanese is argued to have inherently cumulative common noun denotations. To solve this, I propose two possible analyses. The first option is to assume the stratified measurement reference (Champollion 2017) and the second option is to assume that Japanese distinguishes singular count nouns, plural nouns and mass nouns in its syntax (Watanabe 2006; 2017). I discuss the implications of these options in light of the previous literature and provide further data which may suggest that Japanese makes an atomicity distinction both in its lexicon and its syntax.