Volume 21, Issue 3 - Special Issue: Collostructions: New Applications and Extensions; Guest Editors: Stefanie Wulff and Stefan Th. Gries

This paper serves as an introduction to CLLT’s special issue celebrating 20 years of collostructional analysis, a corpus-linguistic method developed in the early aughts to quantify the degree of association between words and constructions and between words in one construction. A variety of case studies revisit critical aspects of collostructional studies, apply the method to native and learner data, and to synchronic and diachronic questions, and chart new ground by extending the method in various quantitative ways (e.g., with the inclusion of more and more diverse variables than just words and constructions or with network models and advanced regression approaches).

This study investigates the semantic structures of degree adverb constructions (DAC) in Mandarin through a network-based approach. We first utilized covarying collexeme analysis to determine significant collexeme pairs strongly associated with the DAC. We then incorporated network-based quantitative methods to analyze the emerging schemas from these collexeme pairs. Our approach involved implementing network analyses at two levels. The collexeme-based network consisted of lexical nodes representing degree adverbs and modified heads, while the construction-based network featured collexeme pairs as nodes. The collexeme-based network reveals that degree adverbs develop distinct semantic preferences for attributes within specific semantic fields, primarily capturing subjective and evaluative aspects of human experience. The construction-based network highlights dynamic interactions among these significant collexeme pairs, showcasing semantically interconnected collexeme pairs forming cohesive constructional families. Furthermore, these semantic groupings underscore the crucial role of metaphorical extension in creating larger semantic clusters for higher-level linguistic generalization. The emergence of semantic coherences from both networks reflects the evolving nature of linguistic generalizations, transitioning from low-level pivot schemas in the collexeme-based network to higher-order generalizations in the construction-based network. This study contributes to the field of usage-based grammar by emphasizing the fundamental role of sequentiality and interconnectedness between linguistic units in linguistic generalization.

The dative alternation has been extensively studied in the world’s languages, and the meanings of the verbs participating in the alternation have been shown to play a key role in determining its argument realization options. The present paper presents a multiple distinctive collexeme analysis approach to the dative alternation in Mandarin Chinese, which involves a choice of one of five functionally similar alternants, and it does so by also discussing several ways to improve how this has been done statistically in most previous analyses. Linguistically, we identify the core semantic differences of the five constructions based on which verbs statistically prefer to occur in which pattern, focusing on semantic potential and direction of transfer. Methodologically, this study contributes to the slowly growing body of studies that use collexeme strengths that are not only less related to frequency than the traditional methods (i.e., association is measured in a less diluted way) and that are directional (i.e., we can focus on one direction of association from the verb to the construction).

Starting from the premise that English negative modal contractions constitute partly variable patterns of associations that include both the preceding subject and the following verb infinitive, the study sets out to investigate distributional differences between can’t , shouldn’t , and won’t and their corresponding uncontracted parent forms. Given that some configurations are assumed to correlate with specific modal meanings (e.g. inanimate subjects and stative verbs > ‘epistemic prediction’; first person subjects > ‘(un)willingness’ or ‘commissive modality’), roughly 200,000 trigrams from COCA are submitted to distinctive covarying collexeme analysis in order to uncover if these contractions and their full forms are conventionalized and entrenched differentially enough to merit their separate treatment on both conceptual and methodological grounds. The results point to probabilistic tendencies, suggesting a cline where won’t and can’t appear to be more emancipated from their respective full-form analogue than shouldn’t . Furthermore, the study showcases how collostructional methods can be applied fruitfully to case studies embedded in Schmid’s (Schmid, Hans-Jörg. 2020. The dynamics of the linguistic system: Usage, conventionalization, and entrenchment . Oxford: Oxford University Press) Entrenchment and Conventionalization Model .

This article presents a corpus-based study of the go (a)round Ving - and go (a)round and V -constructions in American English. More specifically, it addresses the possibility of the constructions serving as pragmatic markers of stance through the collocational phenomenon of semantic prosody. It is argued that the notions of internal and external constructional properties from the early days of construction grammar as well as the corpus-linguistic idea of association patterns would be beneficial to usage-based construction grammatical descriptions of phenomena such as semantic prosody. Drawing on a 248,145,425-word portion of the Corpus of Contemporary American English , both simple collexeme analysis and distinctive collexeme analysis are applied to generate output that feeds into semantic-prosodic analysis. Moreover, standard distinctive collexeme analysis and multiple distinctive collexeme analysis are applied at the level of semantic prosodies in the collexemic fields (i.e., distinctive semantic-prosodic analysis), at the level of verbal category colligations (i.e., distinctive colligational analysis), and at the level of speech act functions of usage-events of the two constructions (i.e., distinctive speech act analysis) as a type of trial balloon. The purpose is to expand semantic-prosodic analysis from focusing merely on lexemes to exploring how other linguistic and pragmatic phenomena may be at play.

The paper reports on a study of the usage of smell verbs over the last 200 years by speakers of American English. The focus is on how the expression of smell changes over time and what this reveals about the way speakers conceptualize and assess smells. The study is based on usage data from the COHA (Corpus of Historical American English). They were quantitatively analysed employing the methods of simple and (multiple) distinctive collexeme analysis. The results of our investigations indicate both a general increase over time in the usage of smell-verb constructions and a noticeable diversification of the smell vocabulary used by American English speakers. Moreover, the results of the collexeme analyses reveal more detailed aspects of the types of smell descriptors people use in smell talk. Reflecting what kinds of smell emitters are most typically and especially closely associated with the individual smell-verb constructions at particular times, they are informative about the sources of smells that are salient enough in our culture and (well-)known enough in the speech community to be used as functional smell descriptors and how these may change over time.

In an attempt to identify possible cases of collostructional transfer in the use of the causative construction [X make Y V inf ] by French-speaking learners of English, two types of analyses are combined in this study. First, a contrastive collostructional analysis compares the verbs occurring in the [V inf ] slot of the English construction and its French equivalent, [X faire V inf Y]. Second, a contrastive interlanguage collostructional analysis compares the verbs used in the [V inf ] slot of [X  make Y V inf ] by native speakers of English, French-speaking learners of English and learners of English from other mother tongue backgrounds. The aim is to identify verbs that are more distinctive of [X faire V inf Y] than of [X make Y V inf ] and that are also more likely to be used by French-speaking learners of English than by other populations, as these verbs could be potential cases of collostructional preferences transferred by learners from French to English. The results suggest that learners might transfer verbs expressing a change of state or location and some individual verbs like discover from the French to the English causative construction. Their dispreference for copular verbs (other than be ) could also be the result of transfer effects.

Japanese features a general noun-modifying clause construction (NMCC) with a more versatile range of semantic and pragmatic interpretations than equivalent constructions in other languages. Motivated by the learning challenge NMCCs pose to Japanese as a foreign language (JFL) learners, this article examines speech data from the International Corpus of Japanese as a Second Language (I-JAS) to compare learner use of NMCCs against a large L1 Japanese corpus. Instances of the construction from both corpora were analyzed to identify high-frequency part-of-speech categories and subcategories in the modifying clause predicate and head noun slots. A simple collexeme analysis was then employed to identify strongly attracted and repelled lexical items among those identified in realizations of the construction. Taken together, findings from these analyses revealed an important connection between the semantic weight of head nouns in NMCCs and the idiomaticity of the construction, with learner productions demonstrating a tendency toward heavy head nouns. This study lays the groundwork for future research seeking to explore the NMCC at different levels of granularity and to improve its treatment in JFL pedagogical materials.

In undertaking any collostructional analysis, a researcher must make decisions concerning the properties of words, constructions, and corpora. Each of these crucial aspects of the analysis can be dealt with in alternative ways: words can be investigated as either lemmas or inflected forms; a construction can be characterized in alternative ways (reliance on semantics or syntax or some combination thereof, the span of the construction, etc.); the choice of corpus (or corpora) will be influenced by whether a researcher has an interest in different genres and varieties, whether the study is synchronic or diachronic, etc. I review various ways in which a researcher’s decisions about words, constructions, and corpora are relevant to a corpus-based study of N waiting to happen , referencing throughout the collostructional analysis of this construction by Stefanowitsch and Gries. The approach adopted here can be seen as supplementing Stefanowitsch and Gries’ original collostructional analysis. It illustrates how multifarious the results of a corpus-based study of constructions can be and serves as a reminder that no one corpus-based measure can possibly answer all the questions linguists might reasonably ask about a construction.