Scholars and educators were once concerned that encouraging children to learn more than one language might have adverse cognitive consequences (Darcy, Reference Darcy1953). And for some linguistic capacities (e.g., fluency, vocabulary) this is often true (Bialystok, Craik & Luk, Reference Bialystok, Craik and Luk2012). Unfortunately, such individual costs might discourage governmental policies that are aimed at fostering multi-lingualism, despite its widely acknowledged societal benefits. Peal & Lambert (Reference Peal and Lambert1962) helped overcome this concern and through her “myth-dispelling” efforts and prodigious empirical output, BialystokFootnote ¹ has pushed the pendulum of opinion in the opposite direction.

Valian (Valian), like many other scholars, wants to believe that the answer to the question that is my commentary's title is: “yes.” So do IFootnote ² . There is an exciting narrative here that meshes nicely with adages from both popular culture – “use it or lose it”, “practice makes perfect” – and contemporary cognitive neuroscience – “neurons that fire together wire together,” derived from Hebb's cell-assembly theory (Hebb, Reference Hebb1949; Klein, Reference Klein1999). I became engaged in this narrative during my 2000–2001 sabbatical spent partly at the Rotman Research Institute where my office was next to that of Ellen Bialystok who was also on sabbatical. I attended lab meetings with Ellen, her Rotman colleague, Gus Craik and her students and, because of my interest and expertise in attention, the Simon effect and cognitive inhibition (e.g., Klein & Taylor, Reference Klein, Taylor, Dagenbach and Carr1994), became a collaborator and co-author on the empirical paper (Bialystok, Craik, Klein & Viswanathan, Reference Bialystock, Craik, Klein and Viswanathan2004) that has become seminal for the literature exploring the theme of this special issueFootnote ³ .

In 2010 when he was a first year graduate student in my seminar Human information processing: in the real world, I encouraged Matt Hilchey to write his term paper on “Costs & Benefits of Bilingualism” because, while I had had several opportunities to highlight the “exciting” Bialystok et al. (Reference Bialystock, Craik, Klein and Viswanathan2004) findings on the “benefits of bilingualism for executive processing”Footnote ⁴ , I had not closely followed the literature spawned by that paper.

Hilchey's outstanding term paper was refined and transformed into a major, comprehensive review of the literature on bilingualism and executive control (Hilchey & Klein, Reference Hilchey and Klein2011) which revealed scant evidence for a bilingual inhibitory control advantage (BICAFootnote ⁵ ) but robust evidence, across the lifespan, for a global advantage (bilingual executive processing advantage, BEPA) – measured using performance on congruent trials – so long as the task itself entailed a sufficient level of response conflict.

Recently, when invited to write a chapter for the Cambridge University Press's Handbook of Bilingualism (Hilchey, Saint-Aubin & Klein, in press), we used the opportunity to place the question into a broader historical context and to update our 2011 review by focussing on the many studies that had appeared since that review was written. As in our earlier paper, our graphic meta-analyses focus on two non-linguistic effects, Simon (including spatial Stroop) and flanker, that entail conflict resolution. These tasks are not the only ones that have been conducted to answer the key question and in both papers we discuss findings from other non-linguistic tasks that might also assess executive functioning. Importantly, however, Simon and flanker tasks have been used frequently enough in published papers to support confident empirical generalizations.

As can be seen in Figure 1, although the more recent literature considered as a whole continued to disconfirm BICA, support for BEPA had surprisingly more or less evaporated.

Figure 1.

The BICA and BEPA effects from all the studies reviewed by Hilchey & Klein (Reference Hilchey and Klein2011) and Hilchey, Saint-Aubin & Klein (in press) are presented here as a function of year of publication from 2004–2014. BICA = interference score for monolinguals minus that for bilinguals (interference = incongruent minus congruent). BEPA = congruent RT for monolinguals minus that for bilinguals. For every study in which the same participants were tested more than once (e.g. with 2- and 4-alternative tasks) or one group of monolinguals were compared to 2 groups of bilinguals, the data from the two tasks and two groups of bilinguals (respectively) have been combined. Grey triangles (children aged <14); grey disks (young adults aged 15 to 30); black squares (older adults aged > 30).

Presenting the data by year of publication highlights how anomalous were the results from the seminal paperFootnote ⁶ and reinforces the suggestion (see Paap, Reference Paap2014) that the early (pre-2011) evidence in favour of a BEPA may have been generated by a dual-edged publication bias: for papers whose narrative is consistent with that of an exciting series of studies by a world-famous expert and against papers reporting null results. A recent on-line article about the “replication crisis” in (social) Psychology rings true for the present discussion:

Indeed, in the keynote article (Valian, Reference Valian2014), after reviewing a wide range of tasks assessing various components of executive function, Valian concludes:

These descriptions of the extant corpus of data on the question, Is there a benefit of bilingualism for executive functioning?, reinforce the conclusions drawn by Hilchey and colleagues (Hilchey & Klein, Reference Hilchey and Klein2011; Hilchey et al., in press) which are highlighted in Figure 1. Mindful of Sagan's cautionary note, and in light of the data reviewed here and by Valian (Reference Valian2014), it is recommended that the pendulum of scientific opinion on this question be put into neutral.