ИСТИНА

This paper is the result of fieldwork ‎ in 2013-2014 in Northern provinces along the south-‎ern Caspian coastline (Gilan, Mazandaran and Golestan; Golestan was separated from Ma-‎zandaran in 1997). All local minority languages are in various stages of endangerment, and ‎some of these regional linguemes put pressure on the others. Generally speaking they are a ‎product of migration, cross-cultural processes, as well as attitudes towards preservation of ‎local languages.‎ Mazanderani and Gilaki are two closely related North-Western Caspian Iranian languages ‎with about 2.5 and 3 million speakers, respectively. Each of them is divided into several dia-‎lects. In both varieties, the speech of plainsmen is opposed to the mountaineers, who call their ‎vernacular Galeshi. The Mazanderani language - that was historically called Tabari – also in-‎corporates dialects of Babol, Sari, Amol; Shamerzad; Gorgan, Ziyarat, Nasrabad, Chashme ‎sorkh; Velatru village (Velayat-e Rud) and others. The inhabitants of Shamerzad do not con-‎sider their local language to be a kind of Mazanderani. The inhabitants of Velatru (Velayat-e ‎Rud) call their local language Gilaki, though they converse with Mazanderani speakers with-‎out any difficulty. Gilaki is also split into dialects of bigger settlements (such as Astara, Ban-‎dar Enzeli, Rasht, Lahijan, Langerud, Rudbar, etc) separated by the river Sefid Rud into ‎western and eastern branch. Additionally, there are a number of intermediate speech varieties ‎of Kelardasht (in Chalus County), and Tonekabon (County); administratively both are includ-‎ed in the territory of Mazandaran province.‎ In addition to other similar features, both languages are characterized by:‎ ‎1. Strong influence of Persian at all levels (phonetics, morphology, syntax, vocabulary).‎ ‎2. Considerable number of Russian borrowings (a higher number in vocabulary, and more ‎frequent in usage, than in modern Persian).‎ ‎3. Reverse Status construction (in terms of the sequence of elements), i.e. (ezāfe construc-‎tion), which unlike Persian begins with a marked dependent determiner followed by the head ‎noun.‎ ‎4. Primarily tonal word stress.‎ In order to analyse the acoustic properties of stress, the speech of native speakers (3 Ma-‎zanderani and 4 Gilaki) was recorded. Multivariate ANOVA analysis showed that the defin-‎ing feature of the stressed syllables in both languages is pitch (p<0.001). Mazanderani stress ‎has no other significant factors, while in Gilaki intensity (p=0.006) and complex 3D-factor ‎‎(pitch-intensity-duration) (p=0.034) are also very important. Thus, the stress in Gilaki is com-‎plex: tonal-dynamic. The main reason for the common basis of the stress (pitch) could be that ‎the speakers of both languages are bilingual, and Persian, being common for them, has tonal ‎stress.‎ In general, Mazanderani and Gilaki speakers can understand each other, though customar-‎ily they communicate in Persian. From a sociolinguistic point of view, Mazanderani is not ‎treated as a prestigious language. Unacquainted people in the streets usually address each ‎other in Persian and change to their local language only after they get acquainted. For in-‎stance, a local passer-by would buy a newspaper from a local newspaper seller, using Persian. ‎Even a local woman would discuss her divorce suit with a local attorney exclusively in Per-‎sian. Parents, especially in rural communities, encourage their children to speak Persian al-‎ways, in order to make their speech undistinguishable from that of residents in the capital. ‎They do this so that their children can seek a better career in Tehran. This striving for a better ‎life and higher social status is one of the reasons why the social basis for Caspian languages is ‎shrinking.‎ We have detected some discrepancies with the generally accepted description of Ma-‎zanderani, especially in the field of phonetics. Although nowadays Mazanderani has no writ-‎ten presence, some books in Arabic script with Latin transcription (dictionaries, collections of ‎proverbs and sayings, legends and narratives) can be found. The actual pronunciation and ‎phonemic composition of some words differ, first, from their presentation in traditional litera-‎ture, and, secondly, from the manner linguistically educated native speakers write themselves.‎ Well-educated Gilaki speakers use Persian alphabet to write books and articles. A well-‎known Gilaki writer M.P. Jektaji publishes the literary journal Gilavā. Here, a special super-‎script letter (‎‏7‏‎) is used to denote the vowel [ə] (ezafe) in the texts. Similarly, some slight vari-‎ations are apparent when we compare the transcription of the Mazandarani magazine Barfo-‎rush with that of the two Mazandarani dictionaries of proverbs (1996; 1990). These demon-‎strate that the written expression is not uniformly consistent. ‎ A corpus of texts on these speech varieties was compiled, and can be accessed on the site ‎of the Institute of Linguistics of the Russian Academy of Sciences. The database is given for ‎these language varieties in audio- and written transcription (in Iranian and IPA variants), Rus-‎sian translation and glossing. For the Gilaki language, we also provide the audio recording of ‎texts from V.S. Rastorgueva, A.A. Kerimova, A.K. Mamedzade, L.A. Pireyko, D.I. Edel-‎man, Gilaki Language (1971). Comparing the speech of our language consultants with this ‎book, we have found out that the vocalism in the book tends towards Tajiki. For instance, ‎there is written the expression Či kuni? “What are you doing?” (text 1). Our consultants pro-‎nounce it Če koni? In many other texts we see the same kind of substitution (e↔i, o↔u). ‎Thus, the phonetic description of Gilaki requires further work and refining. For the Mazanda-‎rani language, we offer recordings of the Mazandarani proverbs and parables (Tayar Yazdan-‎Panah-Lamuki, Dictionary of Mazandarani proverbs, 1996; Mostafa Ansari, Dictionary of ‎Mazandarani proverbs, 1990), as well as some narratives on the lifestyle history and cultural ‎traditions of the people taken during our fieldwork.‎