Mathematics is typically regarded as a universal language, but the choice of language we use to frame arithmetic could recruit different parts of our brain to crunch numbers.
A recent study on students fluent in both French and German has revealed different parts of the brain are used to solve identical problems, depending on the language they're wrapped in.
The research conducted by scientists from the Cognitive Science and Assessment Institute (COSA) at the University of Luxembourg was based on a group of 21 students who all spoke Luxembourgish – a German dialect – as their native language.
Thanks to their education, all of the students had also developed fluency in French, making them proficiently bilingual and perfect test-subjects to answer questions on how language might affect cognition.
Earlier studies by other researchers have argued that the language we use to communicate and think about numerical tasks influences our approach, which raised an interesting question; how, exactly, do people fluent in multiple languages use their brains to do mental arithmetic?
This isn't some trivial curiosity, either. Just under half of the world's population speaks more than one language, which more than just personally useful, provides significant advantages for their wider community.
Along with a number of animals, humans use a process called subitising to rapidly judge small quantities, typically less than five. In fact, this quick-counting process is something young children do before they learn to count.
For more complex evaluations we learn to enumerate and apply logic, which forms the basis of arithmetic.
The researchers set out to determine which parts of the brain were active as arithmetic steadily became more complex, and whether these sections differed much when the same level of problem was presented in another language.
Numbers can be communicated in three basic formats; abstract 'amodal' terms that represent magnitudes; words that describe quantities; and symbolically.
Different sections of the brain are used to deal with quantities that are framed in these different ways.
Neuroimaging research indicates amodal codes tend to be interpreted by the intraparietal sulcus, for example, while verbal processing relies on other parietal regions that aren't specifically concerned with numbers.
To study the different parts of the brain being used to calculate problems of varying difficulty in this experiment, the researchers used functional magnetic resonance imaging (fMRI).
They then asked the subjects to perform addition problems of two grades of difficulty – one relatively simple, involving a pair of single digits; another more complex, based on double-digit sums – while having their brains scanned.
In one trial, they were asked the problems in French. In another, in German.
There was no clear difference in how the students managed the simple arithmetic in either language, at least from the outside.
When the sums got tougher, they took slightly longer to process the information when it was delivered in French, compared with the same computation in German. The volunteers also made more errors when it was delivered in French.
The scans showed a slightly different story, however.
Regions in the temporal lobe generally associated with speech were seen to be slightly more active when simple arithmetic was presented in German, compared with the French.
When given more complex arithmetic, things got a lot more exciting.
In a reversal of the previous results, several areas across the brain appeared to be significantly more active when the problems were presented in French, remaining quiet when communicated in German.
The researchers interpreted the results to suggest the brain was bringing in additional visuo-spatial processing power when the double-digit sums were given in French.
None of this, according to the study, was due to the students mentally interpreting the French.
"Contrarily to previous studies on arithmetic in bilinguals, the current results do not indicate translation processes from French to German," the researchers write.
Instead the subjects could have been summoning mental images of symbols, perhaps visualising the numerals, in order to compensate for differences between the languages.
As populations migrate across the globe as a result of war, economics, or even climate, it'll be increasingly important for teachers and public servants to understand the challenges involved in communicating across language barriers.
Even among those who seem fluent in a language, there's a lot more going on upstairs than first meets the eye.
This research was published in Neuropsychologia.