The Meaning of Verbless Sentences: A Contrastive Corpus Approach

Where do verbless sentences stand in terms of propositional meaning? This paper explores the theoretically controversial phenomenon of the verbless sentence, i.e. structures in which the typical syntactic marker of a sentence – the verbal predicate – is absent. Although such structures exist in many languages (e.g. Bertinetto et al., forthcoming; Janda et al., 2020; Goldberg & Perek, 2019; Bîlbîie, 2017; Landolfi et al., 2010; Merle, 2009; Behr et al., 2005; Guillemin-Flescher, 2005; Fernández & Ginzburg, 2002; Lefeuvre, 2001; Benveniste, 1971), their study has been limited by challenges in automatic retrieval. Focusing on the verbless sentence defined as a string of text that is (a) found in context, (b) contains initial and final marking, and (c) excludes all verb forms, e.g. ‘Yellow.’, ‘Genius.’, ‘A beautiful dream.’, ‘Nice hat.’, ‘From France.’, ‘You liar!’, ‘What genius?’, ‘What brave and lofty words!’, ‘Off with her head.’, ‘Cashier to the flower department.’, we develop automatic processing and take a contrastive corpus approach to the phenomenon. Persuaded that cross-linguistic comparison can reveal linguistic constraints hidden from a monolingual perspective, we examine the structures in English and Russian, two languages with profoundly different typological characteristics regarding the verb (e.g. Stassen, 2013; Weiss, 2013; Kopotev, 2007; McShane, 2000). We develop a methodological framework that combines contrastive linguistics with corpus-driven methods and enunciative analysis, with the goal of (a) providing a corpus-based description of the semantico-pragmatic features associated with the absence of the verb in English and Russian and (b) exploring the theoretical implications of the results for linguistic models of the sentence. Targeting both quantitative and qualitative contrastive analysis, we developed a method for accurate automatic retrieval of verbless sentences, as well as their multiple translation correspondences, and built a 1,4-million-word parallel-and-comparable corpus of 19^th–21^st century translated realist fiction (Bondarenko, 2019; 2021) which we explore in terms of syntactic, semantic and pragmatic features (including semantic classification of statistically characteristic elements and n-grams, verbs implicated in translation correspondences, syntactic ellipsis, information structure).

The traditional view of sentence meaning, i.e. that verbs project an argument structure and thereby specify the meaning of a sentence, is unable to account for propositional meaning in verbless structures without reconstructing the verb either syntactically or semantically. In the present paper we present quantitative contrastive corpus evidence against reconstruction, thereby supporting arguments that verbless structures are neither syntactic nor semantic reductions (e.g. Elugardo & Stainton, 2005; Barton & Progovac, 2005). We also observe the potential of verbless structures to constitute not only constructions but also constructs (in the sense of Goldberg & Casenhiser, 2006: 348-349), i.e. with propositional meaning that is not fully conventionalized, which further complicates their situation (Bondarenko, 2021: 133-141). We propose an explanation of the propositional meaning of verbless structures, that pertains to both conventionalized and non-conventionalized instances, in order to defend their sentential status. 

Relying on contrastive corpus data, we explore the pragmatic implicature of the absence of the verb in English and Russian from a quantitative perspective. Results reveal the mutual use of the structures to mark (dis)agreement, quantification, (in)formality, deixis, questioning and emphasis. Contrastively, English tends to use verbless structures to mark particularly indefinite reference (e.g. (a) What a strange meeting on a strange night.), while emphasis of intensity is statistically important particularly for Russian (e.g. (b) О, эта ужасная тирания большинства!; Oh, this terrible tyranny of majority!). Key n-grams ‘why not’ and ‘what for’ in English suggest two potential conventionalized forms of verbless sentences. Providing a portrait of the constituents of verbless sentences in terms of statistically significant lexical and grammatical elements, our results also identify semantico-pragmatic differences in the use of the structures in the two languages.