Sceptics and believers – or, how to find a path through confounding variables in bilingualism research

Dr Thomas Bak Thomas H Bak is a reader in Human Cognitive Neuroscience at the University of Edinburgh. In addition to his work with Bilingualism Matters, he is a member of the Centre for Cognitive Ageing & Cognitive Epidemiology (CCACE) and the Centre for Clinical Brain Sciences (CCBS).

Parents often tend to be impressed by their children and I am certainly no exception. Today at the breakfast table my wife asked my 3-year old daughter what is in the spotty bag she was holding in her hands. My daughter’s answer was: “I am not entirely sure”. This made me speechless: not only because of the rather fancy word “entirely”, but also because suddenly I realised that this short sentence expresses something that I have been missing a lot in the recent “bilingualism debate”.

As explained in my previous blog post, the attitude towards bilingualism has changed dramatically over the last 50 years. It has moved all the way from a dark force destroying people’s mind to a panacea, making people bright, creative and immune against dementia. It was only a matter of time until the pendulum started swinging the other way again. And indeed, since 2013 we can observe a rising wave of articles questioning the existence of any cognitive difference between mono- and bilinguals.

This in itself could be a positive development. At least some of the claims made about the “bilingual advantage” (not necessarily by the scientists themselves) have been exaggerated and needed to be put into perspective. Moreover, science generally needs debates, questioning of the existing evidence, competing theories. But what we are witnessing at the moment is starting to look less and less like a constructive dialogue based on healthy scepticism and more and more like the emergence of a new fundamentalism: an unshaken belief that any evidence showing positive effects of bilingualism is bound to be wrong while any study failing to find such a difference is necessarily beyond any reasonable doubt. It looks worryingly like a replacement of one exaggeration by another.

It is slightly ironic that some of the proponents of such a view like to refer to themselves as “sceptics” while describing those with diverging opinions as “believer”. Real scepticism is not about questioning other people’s beliefs; it is about questioning one’s own. It is the fundamentalism that assumes that one’s own beliefs are the only correct ones, so any other set of beliefs must necessarily be wrong. Also, a blind belief in non-existence of a phenomenon is not any more sceptical than a blind belief in its existence. A true sceptic will examine the evidence in both directions.

It is widely recognised that bilingualism interacts with a whole host of other variables, from immigration, education and socioeconomic status to individual differences in intelligence, personality and lifestyle. These factors are often referred to as “confounding variables”. For me, they are a fascinating challenge, as reflected in my most recent paper, entitled “The impact of bilingualism on cognitive ageing and dementia: finding a path through a forest of confounding variables” (see the link below). Finding a path can be a scary experience but out of hundreds of mountain walks those I love to recall most are those in which finding the way was particularly demanding, due to snow, fog or difficult terrain.

The path through the bilingual research is certainly challenging and the emerging picture is, unsurprisingly, complex. But a thorough look thorough the existing literature can yield many insights. Firstly, there is overwhelming evidence, coming from the study of different populations (increasingly also beyond the Western World) and using a wide range of methodologies suggesting that bilingualism does influence cognitive functions, across the lifespan as well as in brain diseases such as dementia or stroke. Secondly, both in health and disease, the cognitive effects of bilingualism show a characteristic, remarkably selective pattern. Bilinguals tend to perform better than monolinguals on tests of so-called executive functions, such as inhibition or attention switching. In contrast, they are slower than monolinguals on some language tasks such as picture naming. Thirdly, although various combinations of “confounding variables” occur in most studies, they are by no means more frequent in those showing positive results. In fact, the majority of studies reporting null results are either confounded by immigration or by a suboptimal matching of mono- and bilingual groups. Finally, the so-called publication bias (a tendency to report positive rather than null results) is, if anything, less pronounced in bilingualism research than in many other areas of science. So there is no reason to doubt the results of bilingualism studies any more than we should doubt results of any scientific research. In science, as my daughter would put it, we cannot be entirely sure.

What might be underlying the whole debate is a fundamental misconception about science itself (this also occurs in many others contexts, such as the discussions about the global warming). Many years ago, just after my arrival in Edinburgh, I attended a discussion about enlightenment. A professor of geology, whose name unfortunately I cannot recall said that a big problem in the public perception of science is that most non-scientists think that science is about certainty whereas scientists themselves think that it is about uncertainty. Uncertainty is not a pleasant state to be in and can induce desperate attempts to turn it into certainty. If we cannot be certain that something is the case, isn’t it better to be certain that the opposite is the case? If we cannot prove a large and unequivocal bilingualism effect under all circumstances, isn’t it better to declare once and for all that it does not exist at all?

It would be tempting to associate this need for ultimate certainty with religion, but that would be a very simplistic and distorted view of religion. In reality, we can find a lot of critical thinking and scepticism in major world religions, whether in the work of Nicolaus Cusanus or Moses Maimonides, the Kalam schools of Islam or the “fourteen unanswerable questions” of Buddhism. And the vedaic hymn of creation, at the heart of Hinduism, ends famously with the words: “Whence all creation had its origin, he, whether he fashioned it or whether he did not, he, who surveys it all from highest heaven, he knows – or maybe even he does not know”.

The real opposition is not between sceptics and believers or between science and religion: it is between scepticism and fundamentalism. I cannot know whether my daughter will become one day a scientist or a theologian, a religious person, an agnostic or an atheist. I cannot know what she will believe in. But whatever she does, I hope she retains her healthy scepticism.

More information:

Bak, T. H. (2015). The impact of bilingualism on cognitive aging and dementia: finding a path through a forest of confounding variables. Linguistic Approaches to Bilingualism, Special Issue on Bilingualism and Ageing.
Read an open access copy of the article here