We have known at least since Loftus & Palmer’s 1974 study that language affects witness memory. What is more controversial is that a particular language can affect memory in a language-specific way. This Whorfian possibility has been investigated in different experiential domains and an increasing number of studies confirm that it is not a matter of whether language-specific effects exist but rather when they may be detected and how strong they can be on different occasions.
In this talk I show how speaking more than one language affects memory for events in a universal cognitive domain – motion. When it comes to testing witness memory, verbalisation is inherently present, either actively or tacitly. Witnesses are thinking-for-speaking and thinking-for remembering (Slobin 1987, 1996, 2003) using specific patterns of their language(s). For instance, an English speaker can say ‘The man dropped a bag’ without having to specify whether this was done on purpose or not. In contrast, a Spanish speaker has to make a choice between two different structures – one with intentional and one with non-intentional meaning. So what happens when a bilingual speaks two languages of different types? Is one or both used to encode information in memory and with what impact?
I present experiments on recognition and recall memory, featuring different types of English-Spanish bilinguals (early, late L2 English, late L2 Spanish) and I explain why a bilingual advantage for memory of events seems to be there sometimes but not always (Filipović, 2019). I demonstrate how language typology, the specific type of bilingualism (early vs. late) and the type of communicative situation (single vs. dual language activation) jointly lead to varied yet predictable linguistic and memory outputs.