A look at visual perception constructions
Our study focuses on formal and functional properties of ‘support verb’ complex predicates to conceptualize visual perception, available in Portuguese (both Brazilian and European Portuguese, BP and EP) and in French. Below are two examples, with their (literal) translations:
(1) Vale a pena dar uma olhada em todas as opções.
It’s worth to give a look at all the options.
[Sketch Engine, Portuguese Web 2011 (ptTenTen11)]
(2) Je donne un coup d’œil sur ma montre: 5 minutes 30 sont passées
I give a hit of eye at my watch: 5 minutes 30 have passed
[Sketch Engine, French Web 2017 (frTenTen17)]
Apart from the light support verb donner/dar ‘give’, both languages also allow, among other verbs, lancer/lançar ‘throw’. Such parallells raise the question whether we are dealing with a ‘diasystemic’ construction in HÖDER et al’s (2020) sense, that is, a construction shared by two different languages. This would be in keeping with the fact that Portuguese and French are genealogically and typologically related languages.
To address this question, we analyze data collected via Sketch Engine (Portuguese Web 2011 and French Web 2017). For each language, 1000 candidate occurrences of support verb constructions were extracted, yielding 969 occurrences for Portuguese and 750 occurrences for French, respectively, after eliminating noise examples. Using covarying collexeme analysis (GRIES & STEFANOWITSCH, 2004; FLACH, 2021), we measure degrees of attraction or repulsion between three elements of the construction, the verb, the noun and, if present, the adjective. We thus observe that Portuguese dar occurs with four different nouns, namely olhada, olhadinha, olhadela and vista de olhos, the latter of which is compatible with almost all support verbs. By contrast, French donner is compatible with un regard, une œillade and un coup d’œil, but is not compatible with un œil, while this latter noun is found only with six of the twenty eight verbs retrieved. Drawing on concepts from Cognitive Construction Grammar (GOLDBERG, 1995, 2006; TRAUGOTT & TROUSDALE, 2013), we conclude from these results that a visual-perception support verb diaconstruction exists only at the most schematic ‘macrolevel’, not at the lower ‘meso-level’ or ‘micro-level’. At the lowest level, we find entrenched combinations, which are different, moreover, in the two investigated regional varieties of Portuguese: dar uma olhadela and dar uma vista de olhos are lexically stored sequences in EP only and dar uma olhada and dar uma olhadinha are typical of BP only. We present a network of support verb complex predicates used for visual perception in both languages, showing the place of stored exemplars and more general schemas.
We complement the quantitative analysis with a qualitative one in which we consider the politeness strategies that come with using these expressions to request the addressee to take a look at something, among other speech-act possibilities.