ABB, a salient prototype of collocate-ideophone constructions in Mandarin Chinese

ABB words in Chinese, e.g., hēi-qīqī ‘pitch black’, have been studied for a long time. Most traditional studies analyze these words through derivational rules involving empty suffixes. However, this is problematic, as they are better seen as compounds involving a prosaic A and an ideophonic BB part. By treating ABB as a schema sanctioned by collocate–ideophonic constructions, it is possible to investigate other similar patterns. A corpus study (more than 5,000 tokens) revealed that on the level of schemas, ABB truly acts as a prototype of such constructions, but that it is far from the only pattern to be identified. A second corpus-based study on the level of exemplars showed there are different pockets of salience and non-uniformity in the data from four angles: cue validity, frequency, dispersion, and constructional preference. This paper provides evidence that the traditional ABB narrative needs to be complemented with usage-based data, and grapple with the lexical salience effects this brings along for words involving iconicity.

ABB词语，例如‘漆黑’（hēi-qīqī），长期以来一直是研究对象。大多数传统研究通过涉及空后缀的派生规则来分析这些词语。然而，这种方法存在缺陷，因为它们更应被视为由平实的A部分和象声的BB部分构成的复合词。将ABB视为由搭配-象声结构认可的范式，可以进一步探究其他类似模式。一项基于语料库的研究（超过5,000个token）表明，在范式层面，ABB确实充当此类结构的原型，但并非唯一可识别的模式。另一项基于语料库的实证研究在实例层面揭示了数据在四个方面的不同显著性和非一致性：提示有效性、频率、分布和构式偏好。本文提供了证据，表明传统的ABB叙事需要补充基于使用的数据，并应对由此带来的涉及象征性的词汇显著性效应进行深入探讨。