Kurdish Dialect Recognition using 1D CNN

Dialect recognition is one of the most attentive topics in the speech analysis area. Machine learning algorithms have been widely used to identify dialects. In this paper, a model that based on three different 1D Convolutional Neural Network (CNN) structures is developed for Kurdish dialect recognition. This model is evaluated, and CNN structures are compared to each other. The result shows that the proposed model has outperformed the state of the art. The model is evaluated on the experimental data that have been collected by the staff of department of computer science at the University of Halabja. Three dialects are involved in the dataset as the Kurdish language consists of three major dialects, namely Northern Kurdish (Badini variant), Central Kurdish (Sorani variant), and Hawrami. The advantage of the CNN model is not required to concern handcraft as the CNN model is featureless. According to the results, the 1 D CNN method can make predictions with an average accuracy of 95.53% on the Kurdish dialect classification. In this study, a new method is proposed to interpret the closeness of the Kurdish dialects by using a confusion matrix and a non-metric multi-dimensional visualization technique. The outcome demonstrates that it is straightforward to cluster given Kurdish dialects and linearly isolated from the neighboring dialects.

Definiteness Systems and Dialect Classification

In this article I explore how typological approaches can be used to construct novel classification schemes for Arabic dialects, taking the example of definiteness as a case study. Definiteness in Arabic has traditionally been envisioned as an essentially binary system, wherein definite substantives are marked with a reflex of the article al- and indefinite ones are not. Recent work has complicated this model, framing definiteness instead as a continuum along which speakers can locate referents using a broader range of morphological and syntactic strategies, including not only the article al-, but also reflexes of the demonstrative series and a diverse set of ‘indefinite-specific’ articles found throughout the spoken dialects. I argue that it is possible to describe these strategies with even more precision by modeling them within cross-linguistic frameworks for semantic typology, among them a model known as the ‘Reference Hierarchy,’ which I adopt here. This modeling process allows for classification of dialects not by the presence of shared forms, but rather by parallel typological configurations, even if the forms within them are disparate.

Interrogating the Egypto-Sudanic Arabic Connection

The Arabic dialectology literature repeatedly asserts the existence of a macro-level classificatory relationship binding the Arabic speech varieties of the combined Egypto-Sudanic area. This proposal, though oft-encountered, has not previously been formulated in reference to extensive linguistic criteria, but is instead framed primarily on the nonlinguistic premise of historical demographic and genealogical relationships joining the Arabic-speaking communities of the region. The present contribution provides a linguistically based evaluation of this proposed dialectal grouping, to assess whether the postulated dialectal unity is meaningfully borne out by available language data. Isoglosses from the domains of segmental phonology, phonological processes, pronominal morphology, verbal inflection, and syntax are analyzed across six dialects representing Arabic speech in the region. These are shown to offer minimal support for a unified Egypto-Sudanic dialect classification, but instead to indicate a significant north–south differentiation within the sample—a finding further qualified via application of the novel method of Historical Glottometry developed by François and Kalyan. The investigation concludes with reflection on the implications of these results on the understandings of the correspondence between linguistic and human genealogical relationships in the history of Arabic and in dialectological practice more broadly.

Basic vocabulary of the Karelian and Vepsian languages in the linguistic and geographical aspect

Introduction: Karelian and Vepsian vocabulary has been collected and studied by linguists from Russia and Finland for two centuries. An invaluable source for research in the dialectology of the North-East group of the Baltic-Finnish languages is the «Comparative and Onomasiological Dictionary of the Karelian, Vepsian and Sami Languages» (2007). The dictionary was prepared by staff of the Institute of Linguistics, Literature and History of the Karelian Research Centre of the Russian Academy of Sciences using field research data from 1979–1981. The article reports the main results of applying the statistical method of cluster analysis to the dictionary entries. Objective: the analysis of the basic vocabulary of the dialects of the Karelian and Vepsian languages in the linguistic and geographical aspect using the statistical method of cluster analysis (dialectometry method). Research materials: pre-encoded for being uploaded to the clustering software database lexical data from the «Comparative and Onomasiological Dictionary of the Karelian, Vepsian and Sami Languages» (about 43 thousand units). Results and novelty of the research: the scientific novelty of the research is the application of the statistical method of cluster analysis to large volumes of pre-encoded lexical dialect material. The results of the calculation confirm the conclusions made by linguists earlier regarding the unity of the Vepsian and Karelian languages, as well as the presence of a clear border between them. The question of determination of the linguistic status of the Ludic dialects, which has been the subject of discussions among Russian and Finnish linguists for decades, is resolved in favor of the Karelian dialect on the basis of the material involved in the analysis. The boundaries between clusters outlined by the clustering program for the Vepsian language coincided with its dialect classification. On the Karelian part of the final map, the main bundle of isoglosses shifted north of the border between the dialects of the language, which indicates a more mobile character of its lexical level. The results presented in the article and the method of obtaining them will be later used to develop a linguistically grounded classification of Karelian language dialects.

The classification of the Japonic languages

The chapter starts with an overview of the history of dialect classification in Japan. A puzzling aspect of the distribution pattern of the Japanese dialects is the fact that many features, which cannot all be explained as retentions or simplifications, recur in geographically distant areas. These similarities have been commonly but unsatisfyingly regarded as the result of parallel independent developments. Phonological (including tonal), morphological, and lexical features are selected to illustrate the splits that result in the different branches of Japonic. Based on shared innovations, the new classification at the end of the chapter proposes a Izumo-Tōhoku branch, as well as a Kyūshū-Ryūkyū branch.