نظام‌های‌ برچسب‌گذاری نوایی: مقایسۀ نظام «نواخت‌ها و فاصله‌نماها: توبی» و «ضرباهنگ و زیرو بمی» در زبان فارسی

پژوهش حاضر با هدف انتخاب نظام برچسب‌‌گذاری مناسب برای استفاده در پژوهش‌های ادراکی نوای گفتار، به بررسی ویژگی‌های نظام نواخت‌ها و فاصله‌نماها (توبی)، مشکلات آن و یکی از نظام‌های جایگزین آن پرداخته‌است. در این مسیر، کاستی‌هایِ توبی که با توجه به ماهیت و کاربردهای گستردۀ آن در زبان‌های مختلف مشخص شده‌اند جمع‌بندی و ارائه شده‌است. با مرور آن دسته از پژوهش‌های انجام‌شده دربارۀ آهنگ گفتار فارسی که از نظام برچسب‌گذاری توبی و نظریۀ خودواحد-وزنی بهره گرفته‌اند به نظر می‌رسد مشکلات جهانی این نظام در تحلیل الگوی آهنگ زبان فارسی نیز مشاهده می‌شود. در ادامه نظریۀ خودواحد-وزنی تقویت‌شده و نظام برچسب‌گذاری بر مبنای آن (ضرباهنگ و زیروبمی: رَپ) معرفی شد. سپس برچسب‌های توبی ونظام ضرباهنگ و زیروبمی (رَپ) برای یک نمونۀ فارسی مقایسه و کارایی نظام برچسب‌گذاری ضرباهنگ و زیروبمی برای استفاده در پژوهش‌های ادراکی تایید شد.

an overview of prosodic transcription systems: a comparison of the tones and break indices: tobi and rhythm and pitch: rap

the selection of the appropriate labeling system in any prosodic study depends on the research purpose. in the current research, we have reviewed the labeling system known as tones and break indices (tobi) (pierrehumbert hirschberg, 1990) and its alternative labeling systems including rhythm and pitch (rap) (dilley, 2005; dilley brown, 2005; dilley et al., 2006). the problems of the tobi system were summarized and presented. furthermore, a review of the studies conducted on intonation in persian using the tobi system within the framework of auto-segmental metrical theory (am) showed that the global problems of this system is also observable in its application for the analysis of persian intonation patterns (e.g. eslami, 2005; sadeghi, 2018).originally, the main goal of tobi was to provide a standard transcription tool for labeling intonational features, including prominence patterns and prosodic structure of an utterance so that different users with different working fields could use and interpret each other’s linguistic data. in the tobi transcription system, l and h represent low and high tones, respectively. the diacritic * represents pitch accent, and % represents boundary tones (beckman elam, 1997). this system was initially designed for transcribing the intonation and prosodic structure of english utterances (silverman et al., 1992; beckman hirschberg, 1994; beckman ayers elam, 1997; beckman, hirschberg, and shattuck- hufnagel 2005), as well as a few typologically different languages— for example, gtobi for german (grice benzmüller, 1995), k- tobi for korean (beckman jun, 1996; jun 2000), and j_tobi for japanese (venditti, 1997), and persian (eslami,2005). jun (2022), ladd (2022), and dilley and breen (2022) have identified the shortcomings and problems of the tobi phonetic labeling system to create an international prosodic alphabet (ipra) (hualde prieto, 2016).the rhythm and pitch (rap) system based on enhanced auto-segmental metrical theory (am+) was proposed by dilley and her colleagues (dilley, 2005; dilley brown, 2005; dilley et al., 2006; dilley breen, 2012) to overcome the difficulties of tobi in showing variations and gradation of the categories and to emphasize the importance of distinguishing rhythmic or metrical prominence from pitch prominence. in this system, pitch information is labeled as three tonal targets (h, l, e) and compared to the previous pitch pattern (higher, lower, or equal to it) in the speech signal. therefore, labels in rap have a phonetic representation. metrical prominence (at three levels of strong, weak, and none) and prosodic structure (at two levels, intonational phrase (ip) and intermediate phrase (ip)) are labeled in the rhythm layer. although rap was presented as a “method of transcribing rhythm and related pitch in english” (dilley and brown, 2005 p, 2), the concepts and principles of this system can be applied to other languages.