漢語擬聲 (態) 詞的原型與顯著特徵: 以認知與語料庫語言學方法探討

Prototypicality and salience of Chinese ideophones: A cognitive and corpus linguistics approach

司馬智(Van Hoey Thomas)

指導教授 ： 呂佳蓉

國立臺灣大學/文學院/語言學研究所/博士(2020年)

https://doi.org/10.6342/NTU202003830

摘要

本論文旨在研究漢語擬聲(態)詞的原型與顯著特徵： 以統一一些曾經受各自研究的語言現象，如語言學重疊、聯綿詞、象聲詞，漢語含有所謂的擬聲(態)詞(ideophone)。 然而，所謂的擬聲(態)詞詞彙（ideophonic lexicon）的結構並沒有同質性，而有原型性。 本論文採取歷時的觀點及共時語言學的觀點，顧及到資料的多模態性，特別是書面資料，以補充擬聲(態)詞(ideophone)的類型語言學研究，如補充漢語的資料、歷時觀點、書面語的擬聲(態)詞的用法。
本研究之貢獻包含下列幾點：第一，本研究呈現出擬聲(態)詞的資料庫，名為Chinese Ideophone Database（CHIDEOD）， v. 0.9.3的類型頻率為4948筆，包含上古、中古、現代漢語的資料， 以不同的格式存取得如.rds、.xlsx、.csv、R套件與網路應用程序。 第二，本論文利用CHIDEOD資料庫，以四個個案研究研究擬聲(態)詞詞彙的異樣性。
（一）第一個案例研究，如何畫漢語擬聲(態)詞的邊線，以多重對應分析（multiple correspondence analysis）來摸索本類別的結構，結果證實擬聲詞與單詞素的關係最強；擬態詞與疊字的關係最強，但也更明確地表示不同變數之間的關係。後續研究顯示、語料庫資料也包含所呈現的關係。漢語擬聲(態)詞詞彙沒有一個原型中心，反而有兩個：聲音與非聲音。
（二）第二個案例研究「光」的漢語擬聲(態)詞，以歷時原型語義學(diachronic prototype semantics)本研究呈現該語義場的心理空間(mental spaces)、框架(frames)、語義領域(domains)與其意象基模(image schemas)。 「光」語義場的意思形成成群組，其特徵包含多義、互相有關、動態、包含原型中心的群組。
（三）第三個案例研究先建立資料的向量空間模型(semantic vector space model)，接著摸索表「光」語義場的擬聲(態)詞，以三個顯著性觀點分析：語義顯著性(semasiological salience)、命名顯著性(onomasiological salience)與結構顯著性(structural salience)。
（四)第四個案例研究採取構式搭配分析法(collostructional analysis)，呈現擬聲(態)詞與構式之間的吸引、推斥。 本研究亦論及ABB構式的擬聲(態)詞，表示ABB構式是更抽象「搭配擬聲(態)詞」構式的實例。
總而言之，四個案例研究表示漢語的擬聲(態)詞詞彙並沒統一性，依方法與抽象度來看卻有許多顯著性的特徵。因此，本論文提供資料與方法，以便仔細地考漢語擬聲(態)詞的各種特點。

關鍵字

擬聲詞 ； 擬態詞 ； 象聲詞 ； 擬繪詞 ； 聯綿詞 ； 原型理論 ； 詞彙顯著性 ； 中文 ； 認知 語言學

並列摘要

This dissertation explores prototypicality and salience effects of the variation within the Chinese ideophonic lexicon. Chinese is demonstrated to have ideophones, by unifying previously separately studied phenomena such as reduplication, binomes, and onomatopoeia. However, the “ideophonic lexicon” is not homogeneous; rather, it is prototypically structured. This is demonstrated from a synchronic and diachronic perspective, as well as across different modalities, with special attention devoted to the written modality. Thus, this dissertation aims to address the lacunae within the literature on ideophones, in which Chinese is often underrepresented, diachronic perspectives are scarce, and the ideophonic usage of writing is often neglected.
My original contributions to knowledge include (1) the creation of an open-source database of Chinese ideophones and (2) four methodological perspectives that show how the variation of and within this category is structured. The Chinese Ideophone Database (version 0.9.3) collects 4948 unique onomatopoeia and ideophones (mimetics) of modern Mandarin, as well as Middle Chinese and Old Chinese. It follows a framework that can be reused and updated in future research, and is accessible in different formats (.rds, .xlsx, .csv, R package and online app interface). Based on this database and corpus evidence, the variation of the ideophonic lexicon in Chinese is studied, in four case studies, each with its own methodological lens.
The first case study delineates the boundary of (Mandarin) Chinese ideophones as a category and investigates how it is structured. Using Multiple Correspondence Analysis, the interactions between morphological patterns, orthographic motivation and depiction of sensory domain were calculated. These confirm that sound-depicting ideophones (onomatopoeia) correlate mostly with single morphemes, and that the depiction of movement and sound mostly correlates with full reduplication. However, the analysis also shows how strong other correlations between different values of the parameters are. A follow-up application of Multiple Correspondence Analysis finds that these correlations are also found with corpus data. Important in both applications, however, is the fuzzy overlap between correlations, which strongly suggests that the ideophonic lexicon in Chinese has a dual prototypical core.
The second case study investigates the diachronic prototype semantics of Chinese ideophones in the semantic field of . Through manual study, the mental spaces, frames, domains and image schemas for a sample are followed. The meanings form interrelated polysemous clusters, which are dynamic throughout time, with clear prototypical cores that semantically extend over time and can be transient.
The third case study studies lexical variational salience within the field of ideophones from three perspectives by constructing a semantic vector space based on a historical corpus. These perspectives are semasiological salience, onomasiological salience, and structural salience. These three types of salience show that within the semantic field of , Chinese ideophones are not a homogeneous block. Instead they have different features and elements that stand out, depending on one’s perspective, an observation that can be extended to other types of ideophones and the generalizations that can be made about ideophones as a category.
The fourth case study continues the probing of the heterogeneity within the Chinese ideophonic lexicon, by adopting collostructional analysis to study ideophones used in Mandarin Chinese constructions. The association measures obtained through this method show to what degree individual ideophonic items are attracted or repulsed by certain constructions, and that some items also depend on these constructions to even occur. Furthermore, the well-known ABB construction is addressed and is argued to be an instance of a more schematic construction.
The case studies reveal that the Chinese ideophonic lexicon is not homogeneous, and that many different elements of salience can be found, depending on the perspective and the granularity of the analysis. They constitute an important addition to previous research by nuancing certain intuitive truths about the nature of ideophones.

並列關鍵字

ideophones ； onomatopoeia ； mimetics ； prototype theory ； lexical salience ； Chinese ； Cognitive linguistics

參考文獻

Abbot-Smith, Kirsten Michael Tomasello. 2006. Exemplar-learning and schematization in a usage-based account of syntactic acquisition. The Linguistic Review 23(3). doi:10.1515/TLR.2006.011.

Academia Sinica. 1990a. Zhōngyāng yánjiūyuàn Shànggǔ Hànyǔ biāojì yǔliàokù 中央研究院上古漢語標記語料庫 (Academia Sinica tagged corpus of Old Chinese). http://lingcorpus.iis.sinica.edu.tw/ancient/.

Academia Sinica. 1990b. Zhōngyāng yánjiūyuàn Zhōnggǔ Hànyǔ biāojì yǔliàokù 中央研究院中古漢語標記語料庫 (Academia Sinica tagged corpus of Middle Chinese). http://lingcorpus.iis.sinica.edu.tw/middle/.

Academia Sinica. 1990c. Zhōngyāng yánjiūyuàn Jìndài Hànyǔ biāojì yǔliàokù 中央研究院近代漢語標記語料庫 (Academia Sinica tagged corpus of Early Mandarin Chinese). http://lingcorpus.iis.sinica.edu.tw/early/.

Adamska-Sałaciak, Arleta. 2015. Dictionary definitions: Problems and solutions. Studia Linguistica Universitatis Iagellonicae Cracoviensis 129(4). [unknown2].