基于形符的语义向量空间模型视域下英语缩合词语义分类再思考——以X-tel和X-(a/o)holic为例  

Rethinking the Semantic Classification of English Blends with Token-based Semantic Vector Space Model:The Cases of X-tel and X-(a/o)holic

作　　者：孟庆楠 罗卫华[1] Martin Hilpert MENG Qingnan;LUO Weihua;Martin Hilpert(School of Foreign Languages,Dalian Maritime University,Dalian 116026,China;Institute of English Studies,University of Neuchatel,Espace Tilo-Frey 1,CH-2000,Neuchatel,Switzerland)

机构地区：[1]大连海事大学,辽宁省大连市116026 [2]纳沙泰尔大学

出　　处：《外语教学与研究》2025年第2期190-203,共14页Foreign Language Teaching and Research

基　　金：中央高校基本科研业务费青年教师科技创新项目“基于语义向量空间模型的英语缩合词研究”(3132025331)的阶段性成果。

摘　　要：本研究采用基于形符的语义向量空间模型,以X-tel和X-(a/o)holic型缩合词为例,借助COCA、enTenTen21语料库和R软件,重新审视Plag(2003)关于英语缩合词的语义分类方式。结果表明:Plag的二分法有些过于绝对,其所列举的标准缩合词boatel具有与缩略复合词motel相似的语义右向性分布特征,更宜将所有X-tel型缩合词视作一个连续体。就X-(a/o)holic型缩合词而言,依据Plag的分类框架亦无法对shopaholic和chocoholic进行分类,因为二者的语义中心均与来源词1更近。这说明模板词alcoholic的语义凸显度低于hotel,即析取词缀-(a/o)holic相比于词截-tel更加接近派生词缀。我们认为,采用基于原型范畴的渐进式语义描写方式能更准确地反映具有相似形态句法特征的缩合词之间的关联。Taking X-tel and X-(a/o)holic as two examples, this study employs a token-based semantic vector space model to re-examine Plag's(2003) semantic classification of English blends.Using COCA, enTenTen21 corpora, and R software, we establish that Plag's binary classification is somewhat overly rigid. The proper blend boatel displays distributional patterns of semantic right-headedness similar to the abbreviated compound motel, and it is better to take all the X-tel blends as a continuum. For X-(a/o)holic blends, Plag's classification also fails to classify shopaholic and chocoholic, since their semantic heads are both closer to source word 1. This means the modellexeme alcoholic is semantically less salient than hotel, namely the secreted affix-(a/o)holic is closer to the status of derivational affix than the splinter-tel. We argue that a prototype-based gradient approach to semantic descriptions can reflect the relationships among blends with similar morphosyntactic features more precisely.

关 键 词：英语缩合词 语义分类 词截 析取词缀 基于形符的语义向量空间模型 

分 类 号：H314.1[语言文字—英语]