Published online by Cambridge University Press: 27 November 2014
The goal of my keynote article, “Bilingualism and Cognition” (Valian, 2014), was to resolve the inconsistencies in effects of bilingualism on executive functions, whether the individuals were children, young adults, or old people. To summarize (and sharpen) my argument:
1. Especially in children and young adults, benefits of bilingualism for executive functions are not reliable. In old people, there are benefits for executive functions but contradictory results on delay of cognitive impairment, depending on whether studies are retrospective or prospective.
2. All experiences that have benefits for executive functions and aging – and there are many – yield inconsistent effects. Bilingualism is not alone.
3. Three reasons for inconsistencies in bilingualism and other experiences are:
a. Executive function and cognitive reserve are broad cover terms for a variety of mechanisms, most of which are ill-understood. Because we mean different things by ‘executive function’ from one experiment to the next, we can both think we don't have an effect when we do and think we have an effect when we don’t.
b. Tasks are impure: apparently similar tasks measure different aspects of executive function and measure other aspects of cognition as well. Because we lack a good analysis of tasks, we too often do not know what we are measuring. I encourage readers to examine the demos in the supplementary materials of the keynote article to see for themselves what the tasks are like.
c. Individuals engage in many different activities that may be on a par with bilingualism in their benefits.
4. Different types of bilingual experience are unlikely to explain the variability of findings, given the inconsistencies in extant data on varieties of bilingualism.
5. There is a benefit of bilingualism, but bilingualism competes with other sources of benefits. Especially for children and young adults, whose daily lives are full of cognitively enriching and challenging experiences, we should expect variability in effects of being bilingual.
6. The way forward is to focus on underlying mechanisms.
I thank the commentators! They have provided a rich set of suggestions about how to think about the current data and where to go next. I wish I could do justice to the full range of ideas that they have proposed, but time and space constraints prevent that. This work was supported in part by a grant from the National Science Foundation (BCS-0236700).