In my last post, I started to argue that there is a link between bilingualism – the ability to speak a second language to some degree of proficiency – and linguistic creativity. That is, bilinguals like Jack Kerouac and Ernest Hemingway were better writers in English because they could speak at least one other language as well.

My argument was based purely on anecdote – essentially that, if you look at any list of the 100 greatest novels of all time, a large number of the authors represented will be proficient in another language.

However, there is more than just anecdotal evidence to support the claim. Over the last four decades, there has been a huge amount of research on the potential advantages (and disadvantages) of bilingualism on various aspects of cognition. This includes a growing amount of research on the link between bilingualism and creativity.

Before looking at the research it’s worth first defining some terms. As human beings, we have the capacity to perform various mental processes. Creativity, just one of these processes, is actually a fairly complex concept. In everyday life, being “creative” can mean anything from being good at solving maths problems to being handy with a paintbrush.

Among scientists, a commonly accepted definition is that creativity is the ability to come up with novel and useful (or appropriate) ideas. And the production of these ideas is generally assumed to involve two different mental processes. The first process, divergent thinking, involves producing a number of different ideas in response to some question or problem. The second, convergent thinking, involves searching and analysing these ideas to find the most appropriate one(s). So, if I want to complete the sentence “the man screamed like a…”, divergent thinking is coming up with noun phrases like “banshee”, “baby”, and “boy scout”. Convergent thinking, on the other hand, is settling on “chimpanzee on fire”.

When it comes to producing novel ideas, divergent thinking is key. Ernest Hemingway once wrote: “If you are lucky enough to have lived in Paris as a young man, then wherever you go for the rest of your life, it stays with you, for Paris is a moveable feast”. Coining such a memorable metaphor for the City of Light needed both divergent and convergent thinking but, without the former, the latter would have got Hemingway nowhere. As a result, most researchers interested in creativity tend to focus on measuring individuals’ ability to think divergently. To do so, they tend to employ one of a number of standard tests.

In these tests people might be shown random doodles (something like the inkblot test) and asked to write down as many ideas as they can for what they might represent. Or, they might be shown some visual puzzle and asked to come up with creative solutions to it. One of the most common tests, the Torrance Test, evaluates three aspects of divergent thinking based how many ideas individuals produce in a set period of time (“fluency”), how different the ideas are from each other (“flexibility”), and how different they are to ideas that other people come up with (“novelty”).

In the 1970s, in an early study of bilingualism and creativity, a researcher at the University of North Dakota looked at how elementary school students performed on the Torrance Test (Landry, 1973). The first two groups of students, from the second and sixth grades, were at an elementary school that provided a foreign language programme. The second two groups of students attended an elementary school, which didn’t provide any foreign language tuition. All students in the study had English as their first language. Although there was no significant difference found between the two groups of second-grade students, the researcher found that sixth-grade girls who had received bilingual education did perform better in measures of fluency and flexibility.

In a more recent study, a researcher at the University of Haifa compared the creative thinking abilities of bilingual Russian-Hebrew children in both Hebrew-language and dual-language Kindergartens in Israel, against those of monolingual Hebrew-speaking children (Leikin, 2012). The children, from similar socio-economic backgrounds, were tested at the start of Kindergarten and then, again, one year later. The researcher found that there was no significant difference between the divergent thinking of the three sets of students in the first round of tests. However, when the tests were repeated, the bilingual students in the dual-language programme performed significantly better (in terms of flexibility and novelty) compared to the monolingual group, suggesting that regular exposure to two languages at school did enhance creative thinking.

Doing this sort of research is always challenging and it’s almost impossible to control for all the factors – age, socioeconomic background, education level, cultural background, and so on – that might affect the creative performance of the individuals concerned. Even establishing a study group with similar levels of mono- or bilingualism is challenging enough. When it comes to speaking multiple languages, everyone is different, for example in terms of whether they speak a second language at home, when they began learning the language, how many other languages they might speak, and what those languages are. For example, researchers (including an old Professor of mine) recently found that habitual code-switchers performed better on the Torrance Test than non-habitual ones (Kharkhurin & Wei, 2015). Inevitably, whenever you measure the creativity of any random sample of people – bilingual or otherwise – you will get a wide range of scores. Although everyone is creative, some people are just naturally more creative than others.

Despite the research challenges, most researchers are now in agreement that there is a positive correlation between bilingualism and creative thinking. That is, on average, bilinguals are more likely to be more creative thinkers (and therefore more creative writers) than their monolingual counterparts.

The next question, of course, is why?

A few explanations have been offered as to why this might be the case. One theory is simply that bilinguals benefit from a wider range of experiences than monolinguals because they operate in more than one languages and, often, within more than one culture. As a result, they have access to a wider range of ideas, which they can combine together to form new and novel ones.

One of the most appealing theories was suggested by Anatoliy Kharkhurin, a psychologist from the American University of Sharjah, in his book Multilingualism & Creativity. Kharkhurin points to the fact that, within our brains, all our knowledge of concepts and things is stored in a complex, interconnected semantic network. This network is known to have two layers or levels.

At the bottom level are the concepts themselves – things like DOG and CAT and CASTLE. Within this layer, related concepts are horizontally connected. For instance, the concepts DOG and CAT might be connected since both are quadrupeds and common household pets (and, between them, they account for the majority of the videos on Youtube). As a result, thinking about cats is likely to make you think about dogs too. The concepts DOG and CASTLE, on the other hand, are unlikely to be interconnected. So, thinking about dogs won’t immediately make you think about turrets and drawbridges.

In the next level up (the lexical level) are the linguistic labels for each of these concepts – for example, the words “cat”, “dog”, and “castle”. Importantly, the levels are vertically interconnected so that the concept CAT is connected to the lexical item “cat”, DOG is connected to “dog”, and CASTLE is connected to “castle”.

There are also further horizontal connections within the lexical level so that “dog” might be connected to words like “log” and “fog” because they each share two phonemes. As a result, if someone said to you the word “dog”, because of the various links in your semantic network, you might think about cats and logs. But you would be unlikely to think about castles.

That is, unless you also spoke French. Although bilinguals will still have only one set of concepts in the bottom layer of their semantic network, they will have two sets of linguistic labels in the next level up. So a French-English bilingual will have both “château” and “castle” connected to the concept CASTLE, and both “cat” and “chat” connected to the concept CAT. Therefore, if you said the word “dog” to a French-English bilingual it might make them think of cats, which might make them think of the word “chat”, which might make them think of the (phonetically related) word “château”, and therefore the concept CASTLE – a train of thought which just isn’t open to a monolingual speaker of English.

This process of language-mediated concept activation is one of the key processes, Kharkhurin proposes, behind the positive correlation between bilingualism and creativity. And, by way of a real life example of this process in action, here’s a passage Jack Kerouac wrote in his diary in February 1950:

“In my sleep I referred to myself, in French, not as “writer” but as arrangeur – he who arranges matters; at the same time, I associated this fraction with eating supper (manger). I woke up to remember this.”

So there you have it. On average, bilinguals are found to be more creative (linguistically and otherwise) than monolinguals, and there are some convincing theories for why this might be the case.

Of course, there’s far more to good writing than divergent thinking. But it seems fair to say that, if you aspire to be a successful novelist and you already speak a second language, then you’re off to a good start.

References

Kharkhurin, A. V., & Wei, L. (2015). The role of code-switching in bilingual creativity. International Journal of Bilingual Education and Bilingualism, 18(2), 153-169.
Landry, R. G. (1973). The Relationship of Second Language and Verbal Creativity. The Modern Language Journal, 57(3), 110-113.
Leikin, M. (2012). The effect of bilingualism on creativity: Developmental and educational perspectives. International Journal of Bilingualism, 17(4), 431-447.