اطلس گويشي و گويش‌سنجي كرانه جنوب‌شرقي درياي مازندران به روش بسامد واكه

Dialectometry of the Southeastern Caspian Sea Coast via Vowel Frequency Method: A Dialect Atlas

گويش سنجي روشي كمّي براي اندازه گيري تفاوت‌هاي ميان گويش هاي مختلف است كه با محاسبه آماري مقدار اين تمايزات در نمونه هاي گسترده اي از ويژگي هاي زباني در يك منطقه گويشي به دست مي‌آيد. هدف پژوهش حاضر نيز ارايه نمونه‌اي از اطلس گويشي استان مازندران بر مبناي روش‌هاي گويش‌سنجي است. در پژوهش حاضر كه مبتني بر روش توصيفي هم‌زماني و دادگان ميداني است، آوانگاشت معادل‌هاي محلّي 62 واژه مورد تحليل انبوهه قرار گرفت و منطقه‌هاي گويشي كرانه هاي جنوبي درياي مازندران بر اساس بسامد واكه‌ها مكان‌يابي شده است. موادّ زباني متشكل از 73 مصاحبه است. ميانگين سني گويشوران 36 سال، متوسط سواد آنها در حد آموزش ابتدايي و 38% آنها زن بودند. نقشه‌هاي تفسيري حاصل ضمن تعيين چهار منطقه گويشي شمال‌غربي، غربي، شمال‌شرقي، و جنوب‌مركز با آلفاي كرنباخ 84/0، نشان مي‌دهد كه واكه‌هاي پيشين /i/ و /a/، بيشترين شاخص تاثير را در گوناگو‌ني‌هاي منطقه‌اي دارند.

- Introduction Dialect geography is a sub-discipline of linguistics, which investigates the regional distribution of specific linguistic features in a given geographical area and represents them in linguistic maps. A linguistic map is a thematic map that displays geographical distribution of linguistic features, speakers of a language, or isoglosses of a dialect continuum. One of the quantitative methods of dialect geography, which was developed in 1970s, and improved later by the use of computers, is dialectometry. Dialectometry is a quantitative method to measure dialect variations applying statistical calculations and aggregate analysis in large samples of linguistic features for a dialect region. Quantitative methods in dialectometry are called aggregate analysis, because instead of considering individual linguistic feature distribution at once, they measure the differences among several variables altogether, then calculate the relative differences in the given data set. Here, the basic idea is that taking multiple variables into account draws a more accurate picture of the relationship among language varieties. Hence, current research aimed to draw a dialect map of the southeastern Caspian Seacoast, through aggregate analysis of Mazandarani dialect data. 2- Theoretical Framework Following Hoppenbrouwers and Hoppenbrouwersʹs (1988) phone frequency method, current research aimed to draw a dialect map of the southeastern Caspian Seacoast, through aggregate analysis of Mazandarani dialect data, taken from The National Project of the Linguistic Atlas of Iran. Hoppenbrouwers and Hoppenbrouwers (1988) described an experiment in which languages were compared on the basis of phonetic texts from the International Phonetic Association IPA (1949). For each text, the frequencies of phones were obtained. Since not all samples were the same size, relative frequencies were used instead. The distance between two languages was equal to the sum of the absolute values of the differences between the corresponding relative phone frequencies. 3- Methodology The research setting is located between the southeastern Caspian Sea coast and the northern slopes of eastern Alborz mountain ranges. Language materials, i.e., Mazandarani local equivalent of 62 lexical entries (4526 equivalent in total), were collected for every site through questionnaires, accompanied by 73 interviews with NORM (Non-mobile old rural male) speakers (38% female) of 36 years old with an average primary level of education and. For the purpose of data analysis, firstly, local equivalents were transcribed using 38 IPA symbols (29,394 tokens in total). Secondly, vowels were extracted and statistically analyzed. Then, the absolute frequency of each vowel in the database was calculated. Next, the relative frequency of each vowel for every dialect region was calculated and normalized. Associated relative frequencies were mapped to reveal the regional dialect continuum. Finally, in order to account for the resulting dialect regions, every vowel’s impact index -which equals to absolute average value of the relative frequency of each vowel to its absolute frequency in the corpus- in differentiating dialect regions, was calculated. As for the reliability of the research findings Cronbach’s ? was applied. 4- Results and Discussion Normalized relative frequency and impact index of vowels account for regional variations based on vowel system variable, namely: with the impact indexes of -2.34 for the front, closed, spread, long vowel /i/ ,-1.55 for the front, open, spread, short vowel /a/ ,1.52 for the back, half-open, half-rounded, short vowel /? / ,1.50 for the back, closed, rounded, long vowel /u/ ,1.41 for the central, half-open, half-spread, short vowel /? / , 1.06 for the back, open, rounded, long vowel /? / , 1.02 for the central, half-spread, short vowel / ? / , and finally 0.95 for the near back, half-closed, spread, short vowel /? / 5- Conclusion and Suggestions The interpretative maps, based on interactive beam maps, illustrate four regional dialects in northwest, west, southeast, and south-central highlands. Although all four regional dialects share the same vowel system, the relative frequencies of the vowel sound in each region varies according to different phonetic processes, hence, there come different impact indexes. In brief, the front vowels /i/ and /a/,in comparison to central vowels /?/ and /?/, impose uttermost differences in the index on regional variations. Key Words: Linguistic Atlas, Dialectometry, Vowel frequency method, Mazandarani language. References (in Persian) Yadollah, P. (2007).The national project of the linguistic atlas of Iran. Tehran: ICHHTO. Najafian, A., Musavi Myangah, T., Rovshan, B., & Mollaye Pashaye, S. (2014).Dialectometric recognition of Mazandarani language varieties’ linguistic and geographic pattern located off the Gorgan Golf through central Mazandaran. Retrieved from http://lrr.modares.ac.ir/article_11277_0.html (Agust 31,2014) References (in English and Dutch) Hoppenbrouwers, C., & Hoppenbrouwers, G. (1988).De feature frequentie methode en de classicatie van Nederlandsedialecten. TABU: Bulletin VoorTaalwetenschap,18(2),51-92. Hoppenbrouwers, C.,&Hoppenbrouwers, G. (2001).De indeling van de Nederlandsestreektalen.Dialecten van 156 steden en dorpengeklasseerdvolgens de FFM. Koninklijke Van Gorcum B.V., Assen. Heeringa, W. (2004).Measuring dialect pronunciation differences using Levenshtein Distance, Unpublished Doctoral Dissertation). University of Groningen, Groningen. The International Phonetic Association,(1999).Handbook of the international phonetic association: A guide to the use of the international phonetic alphabet. Cambridge: Cambridge University Press. Jain, A. K.,&Dubes, R. C. (1988). Algorithms for clustering data. New Jersey: Prentice Hall. Kruskal, J.B. (1999). An overview of sequence comparison. In D.Sankoff,& J.Kruskal (Eds.),Time warps, string edits, and macro molecules; Thetheory and practice of sequence comparison(pp. 1-44).Stanford: CSLI. Nerbonne, J., & Heeringa, W. (1998). Computationelevergelijking en classificatie van dialecten. Taal en Tongval, TijdschriftvoorDialectologie, 50(2), 164-193. Nerbonne, J., & Heeringa, W. (2001). Computational comparison and classification of dialects. Dialectologia et Geolinguistica: Journal of the International Society for Dialectology and Geolinguistics, 9, 69-83. Osenova, P. & W. Heeringa & J. Nerbonne (2010). “A Quantitative Analysis of Bulgarian Dialect Pronunciation”. In: Zeitschrift für Slavische Philologie, 66 ( 2):425-458. Weijnen, A. (1946). “The borders between the Brabant dialects cross east-north”. In lectures on North Brabant dialect issues. Royal Dutch Academy of Sciences, April 12, 1944, Volume VIII.