English comparative constructions at different levels of schematicity: what is the role of adjective-specific variability?
Date issued
September 26, 2025
In
Cognitive Linguistics
Abstract
<jats:title>Abstract</jats:title>
<jats:p>This study analyzes the English comparative alternation (morphological <jats:italic>prouder</jats:italic> and periphrastic <jats:italic>more proud</jats:italic>) with regard to adjective-specific variability. Substantial empirical evidence indicates that speakers redundantly represent both abstract schemas (e.g., ADJ + <jats:italic>-er</jats:italic>) and their specific instantiations (e.g., <jats:italic>prouder</jats:italic>) as symbolic units, which together constitute a network of constructions at different levels of schematicity. Against this background, we offer a re-analysis of data from (Hilpert, Martin. 2008. The English comparative - Language structure and use. <jats:italic>English Language & Linguistics</jats:italic> 12(3). 395–417), which yields new results on the basis of a mixed-effects (multi-level) model that incorporates individual adjectives as random effects. The results demonstrate robust lexical idiosyncrasies in this alternation. First, many adjectives exhibit tendencies that deviate from the general conditioning factors incorporated as fixed effects. Furthermore, frequency of adjectives in comparative forms positively correlates with the degree of skewness. In other words, the more an adjective is used in comparative forms, the stronger its preference becomes, regardless of the direction of the preference. These findings support the idea that competition between linguistic forms occurs across different levels of schematicity: at the schematic level, the competition between variants is governed by probabilistic factors, while at the lexically specific level, the conventionalization of concrete forms can override these factors.</jats:p>
<jats:p>This study analyzes the English comparative alternation (morphological <jats:italic>prouder</jats:italic> and periphrastic <jats:italic>more proud</jats:italic>) with regard to adjective-specific variability. Substantial empirical evidence indicates that speakers redundantly represent both abstract schemas (e.g., ADJ + <jats:italic>-er</jats:italic>) and their specific instantiations (e.g., <jats:italic>prouder</jats:italic>) as symbolic units, which together constitute a network of constructions at different levels of schematicity. Against this background, we offer a re-analysis of data from (Hilpert, Martin. 2008. The English comparative - Language structure and use. <jats:italic>English Language & Linguistics</jats:italic> 12(3). 395–417), which yields new results on the basis of a mixed-effects (multi-level) model that incorporates individual adjectives as random effects. The results demonstrate robust lexical idiosyncrasies in this alternation. First, many adjectives exhibit tendencies that deviate from the general conditioning factors incorporated as fixed effects. Furthermore, frequency of adjectives in comparative forms positively correlates with the degree of skewness. In other words, the more an adjective is used in comparative forms, the stronger its preference becomes, regardless of the direction of the preference. These findings support the idea that competition between linguistic forms occurs across different levels of schematicity: at the schematic level, the competition between variants is governed by probabilistic factors, while at the lexically specific level, the conventionalization of concrete forms can override these factors.</jats:p>
Publication type
journal article
File(s)
