A Corpus-based Constructional Account of the Chinese bi-marked Comparative Constructions

Abstract

Comparative constructions with the comparative marker bǐ have long been a central topic in Chinese linguistics, as they display a wide range of derivations. While previous studies have yielded fruitful insights, most of them focus on descriptive accounts without adequate explanations with empirical evidence. There remains a need of empirical investigations into the full range of collocational patterns and a principled account of the interrelations among comparative-related constructions. In particular, what mechanisms and motivations underlie these unique derivations, and what kinds of form-meaning correspondences are characteristic of the bǐ comparatives? These questions are to be addressed by the present research, which undertakes a corpus-based quantitative approach to investigate the constructional properties of the bǐ comparative, in pursuit of both empirical and theoretical significance.

The bǐ-marked comparative construction (or bǐ-Comparative) is defined prototypically with four essential components: Target, Marker bǐ, Standard, and Attribute, as illustrated in the example, wǒ_Target bǐ tā_Standard gāo_Attribute 我比他高 ‘I am taller than he’. To fully explore the range of from-meaning mapping variations of the four components, this research aims to focus on three interrelated issues: (i) the nature of Target-Standard mismatch in the comparative schema, as in nǐ bǐ qùnián shòu 你比去年瘦 ‘You are skinnier (this year) than last year'; (ii) the collocational patterns with adverbial modifications, such as the use of degree adverbs, and (iii) the distributional differences among alternative negative comparative constructions, marked with méi, bù or méiyǒu.

To address the three issues, this research adopts the framework of Construction Grammar and applies two complementary methods: the Behavioral Profile (BP) approach and Collostructional Analysis (CA). The BP approach enables fine-grained corpus annotation to distinguish mismatched Target-Standard patterns from matched ones, thereby refining earlier accounts of comparative variation. Collostructional Analysis is employed to examine collocational preferences with regard to degree adverbs and attributive predicates, as well as the distributional patterns of alternative negative comparatives. Together, these methods provide a quantitative, construction-based account for the form-function mapping relations in the comparative domain.

The three issues are examined with three studies that constitute the three main chapters in this dissertation. Study 1 (Chapter 4) investigates the commonly-seen mismatch with the BP approach and reveals that Target-Standard mismatches often correlate with three distinct factors: there is an explicit Dimension of comparison, the attributive predicate is phrasal, or the Target denotes a temporal category. Based on the findings, the bǐ-Comparative is redefined as Comparison-Evaluative Construction (CEC), a member of the constructional family of evaluative expressions. The observed form-meaning mismatch is made possible in Chinese with the topic-prominent structure, in which the Standard marked by bǐ serves to signal comparative evaluation, and the missing Target may be inferred and coerced into the construction under topic prominence, licensed by contextualized metonyms.

Study 2 (Chapter 5) investigates the collocational preference of degree adverbs with Collostructional Analysis and shows that synchronically, the CEC predominantly co-occurs with comparative degree adverbs (e.g., gèng ‘more’), typically positioned after bǐ to mark an increase in degree. Non-comparative adverbs, by contrast, are generally incompatible with the CEC due to semantic redundancy or conflict. Diachronic analysis further reveals that non-comparative degree adverbs were compatible with the construction in the Song and Yuan dynasties, but their uses declined sharply with the constructionalization of the bǐ comparative, nearly disappearing in early Modern Chinese.

Study 3 (Chapter 6) investigates three negative comparative patterns with different negators, bù-bǐ, méi(yǒu)-bǐ, and méi(yǒu), through Collostructional Analysis. The data show that each construction exhibits distinct preferences for attributive predicates, reflecting their semantic and pragmatic differences. In particular, the contrasts between the bù-bǐ construction and the méi(yǒu)-bǐ constructions are linked to the presuppositional functions of negation and to pragmatic shifts in evaluative orientation.

Overall, this study demonstrates that the derivations in the bǐ-marked comparative construction are not irregular deviations, but systematic patterns shaped by Chinese-specific form-function interactions. It also shows that the collocational behavior of degree adverbs and the distribution of alternative negative comparatives pertain to important evaluative orientations in usage. By integrating constructional account with corpus-based empirical methodology, this research helps to identify the correlational factors for the long-debated issue of form-meaning mismatch, detect the distributional preference of varied degree adverbs, and the usage patterns with different negative markers in the comparative constructions. It helps to advance our understanding of the interaction between form, meaning, and historical change in Mandarin comparative expressions. More importantly, it proposes an integrated account of the bǐ-marked Comparative within the family of evaluative constructions in Mandarin, contributing to the theoretical discussion of constructional networks and the dynamic process of constructionalization in natural language.

Date of Award	31 Dec 2025
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Guofu LONG (External Supervisor) & Meichun LIU (Supervisor)