Abstract :
[en] Although quantifiers themselves have been studied thoroughly (e.g. Barwise & Cooper 1981; Gärdenfors 1987; Langacker 1991, 2008, 2016, 2017; Doetjes 1997; Benninger 1999; Radden & Dirven 2007), the modification of quantifiers, i.e. quantity modification, has received scant attention, with the exception of Njende et al. (2015, 2017) and Davidse et al. (2018, fc.). Examples (1) and (2) below illustrate how two relative quantifiers, a subtype of quantifiers comparing a predicated mass P expressed by the quantifier to a reference mass RT (Langacker 1991), undergo quantity modification in English and French respectively, i.e. how the quantity they denote is altered by a modifier highlighting the partial (as in (1) and (2)), full or non-coincidence between P and RT.
(1) Almost all of my friends have been fined either for having no seatbelt, or having no lights on their bikes. (YCCQA_uk)
(2) Beaucoup font aussi la lecture de video, comme Vlc qui peut lire presque tous les formats video existant sans installation de codec. (YCCQA_fr)
‘Many read videos too, such as VLC, which can read almost all existing video formats without installing codecs.’ (Translation mine)
This corpus-based synchronic study investigates quantity modification in translation. More precisely, it aims to inventory the choices made and translation strategies used by translators encountering modified quantifiers.
English all and French tout were selected for this study as they can be modified by a variety of adverbs and are relatively frequent in various types of speech. These two relative quantifiers can pair with proportional modifiers (Njende et al. 2017, building on Paradis 1997, 2000, 2001) to further detail the degree of coincidence between a predicated mass P and a reference mass RT, and can undergo three types of modification: totality modification (3), which flags full coincidence of P and RT (Njende et al. 2017); approximating modification (4), which highlights partial coincidence of P vis-à-vis RT (Njende et al. 2017), and negational modification (5), which stresses the non-coincidence between P and RT by means of a negator serving as a modifier of the quantifier (Marion 2021).
As the French negative construction (ne)…pas can take up different positions in the sentence, it will also prove interesting to study whether negational modification in English is kept in the French translations or whether it is replaced by, for example, sentential negation.
A semantic specification phenomenon (SSP, underlined in (5)) was often attested in the monolingual data when a negational modifier preceded a quantifier. SSPs consist in the addition of an element specifying the semantics of negational modification by means of setting up an explicit contrastive relationship (e.g. by adding many after not all, as in (5)) or by means of exemplifying (e.g. listing possibilities illustrating how P is different from RT). This study will verify whether SSPs also occur in translation and, if so, whether they are equally frequent, and account for the way translators handle SSPs. The results obtained will allow to better understand the vague semantics of negational modifiers and will contribute to expanding the theory on quantity modification.
(3) Absolutely all the expenses (…).1
(4) (…) allergic to pretty much all animal hair.1
(5) Not all people who don't vote for Obama are racist. Actually, many of them aren’t.1
Data were extracted from the sixth version of the Europarl Direct parallel corpus (Cartoni & Meyer 2012), favoured over the seventh version as the former benefits from accurate language tagging. Europarl Direct consists of proceedings from the European parliament and totals approximately 50 million words per language collected from April 1996 to December 2010 (Koehn 2005, Cartoni & Meyer 2012). Provided that sufficient data was available, datasets of 300 instances per language pair were extracted using AntConc (Anthony 2010) as the source files were already aligned, thus making a parallel concordancer unnecessary. On the basis of the data selected, a parallel study contrasting the translations of English all into French and French tout into English was performed.
Preliminary results suggest translators either opt for quite literal, formally corresponding (Catford 1965) translations, as in (6), or resort to a more synthetic translation, as in (7) where the combination almost all is rendered as the compound noun quasi-totalité changing the quantity modification into a type of degree modification.
(6) Virtually all our states (…). (Europarl6_uk)
‘Quasiment tous nos États (…).’ (Europarl6_fr)
(7) (…) for almost all its exports (…). (Europarl6_uk)
‘(…) pour la quasi-totalité de ses exportations (…).’ (Europarl6_fr)
Not only will this study allow to explore quantity modification in translation and chart hitherto unexplored territory, but it will also shed light on the possible translations for both items under scrutiny and the main strategies translators resort to when confronted with the quantity modification of all and tout. A better understanding of how negational modification works and is translated into both languages under study is also targeted in this study.
