The linguistic relativity hypothesis proposes that the language we speak affects the way we perceive the world (Whorf, 1956). This hypothesis is at the origin of a vigorous debate amongst philosophers, linguists, psychologists, and more recently cognitive neuroscientists: Does language shape thought?
In the work presented here, we have chosen to investigate language-time conceptualisation relationships in bilinguals whose two languages refer to time differently. After reviewing key aspects of the literature and summarising recent findings of experiments that have tested linguistic relativity predictions, I present four experimental studies using event-related brain potentials to test cross-language grammatical and lexical influences on time conceptualisation in
highly proficient Chinese-English bilinguals and native English controls.
In Study 1, we set out to characterise how native speakers of Mandarin Chinese –a tenseless language– deal with reference time misalignments conveyed by tense in English. Participants made acceptability judgements of sentences in which the adjunct clause started with the connective ‘after’ and was either temporally aligned or not with the tensed verb of the main clause. Chinese
participants’ data lacked an N400 ERP effect that we had observed in their English native peers, suggesting that they experience difficulties locating events on a timeline when temporal information is conveyed by tense.
In Study 2, we asked Chinese-English bilinguals to pay direct attention to temporal information conveyed by tense in order to test whether they have a different conceptualisation of the recent past when reading in English. For instance, the sentence “*After she has worked in this hospital for ten years, she retired to Spain” features a time clash because an event situated in the past cannot follow a just-completed event. Although bilinguals could detect such clashes just like native English controls, the expected N400 modulation was entirely absent in their data. Despite their mastery of English grammatical rules, Chinese-English bilinguals thus failed to conceptually distinguish past and recent past on the basis of tense.In a further two studies, we turned to vocabulary differences between Chinese and English to see whether specific expressions of Chinese can selectively shape time conceptualisation.
In Study 3, we focussed on Chinese spatiotemporal metaphors that are inconsistent with the future-infront convention of time representation. We tested temporal conceptualisation along the sagittal axis using stimuli –days of the week and years– played through loudspeakers that were physically located
in the front or the back of participants. Brain activity revealed interference when space-time combinations clashed with spatiotemporal metaphors, but, surprisingly, only when participants operated in English, showing that time conceptualisation can selectively affect the second language.
In Study 4, we tested the controversial claim that Chinese organises time along the vertical axis. We adapted the arrow flanker task by replacing the middle arrow with spatiotemporal metaphors for months (e.g., 上个月–‘last month’, literally ‘up month’) and tested whether expected N400 modulations would generalise to purely temporal expressions (e.g., 去年 –‘last year’, literally ‘gone year’). N400 modulations elicited by spatial words (i.e., up / down) and spatiotemporal metaphors surprisingly generalised to non-directional temporal expressions of Chinese, providing strong evidence in favour of the claim that Chinese native speakers conceptualise time along the vertical axis.
Taken together, these results suggest that Chinese second language speakers of English do not process English tense at a semantic level as they read and that temporal expressions idiosyncratic to Chinese have a language-dependent impact on time conceptualisation in bilinguals. We conclude that an intuitive proposition put forward by Whorf in the 1940s may have been correct and possibly had an even stronger basis than he, himself, was prepared to admit.