In my last post I talked about a research study commissioned by the FBI, of all organizations, on what it takes to be a good translator. The study concluded that speaking a second language is not a predictor of translation ability and that proficient translators possess a separate skill which the researchers literally referred to as an “X-factor”. Based on this discovery they concluded that candidate translators should be given actual translation tests and not simply be tested on language comprehension.

That got me thinking. On the one hand, we can all breathe a sigh of relief that the FBI, which relies on accurate information to make life-or-death decisions, has taken a giant step forward in making sure that their translators are actually up to the task. But why commit all this time and money to a study that is pretty much the equivalent of kicking in an open door?

I think the issue is not so much that the FBI was hugely surprised at this finding; in fact, I’m pretty sure they were not. I think the underlying problem they have been trying to solve has more to do with time, objectivity, and staffing. A multiple-choice vocabulary and reading comprehension test doesn’t take a lot of time to grade, the answers are objectively right or wrong, and anyone with the answer sheet can grade the test. An actual translation test takes a lot more time to assess, there are usually multiple ways to translate a sentence correctly, and the reviewer therefore needs to be a translator him/herself. And even then, as every translator knows, it’s notoriously hard for editors keep a firm handle on their own biases and not confuse subjective preference with objective error. It makes sense that an organization whose primary business is not translation would hope that a simple, straightforward language test would be an adequate way to test translation skills, and it speaks for them that they finally concluded that this was not the case.

But I think they stopped too soon. What they should be asking themselves is: what is this X-factor, exactly? And what if there was an objective way to test for this X-factor as a predictor of translation ability?

As it happens, a lot of research has been done on brain activity since the development of functional Magnetic Resonance Imaging (fMRI), and researchers have been able to pinpoint exact areas of the brain that are activated during specific activities. As it further happens, several of these studies focused specifically on what happens in the brain during translation activity.

The first item of interest is that this imaging research confirms the conclusions of the FBI study, as stated for example in “Explore the Brain Activity During Translation and Interpreting Using Functional Near-Infrared Spectroscopy” by F.M. Lu et al.: “Interestingly it was found from previous work that the findings using the native language (L1) or the non-native language (L2) in different settings during monolingual communication is not the same as that from translating […] It follows that translation-specific processes cannot be directly inferred from research on, or models of the bilingual brain.”

Second, there are certain areas of the brain that are activated specifically during translation. Keerthi Ramanujan et al. concluded in “An fMRI Study of Executive Control During Translation and Oral Reading in Cantonese-English Bilingual Speakers” that “when translation and oral reading were contrasted, greater activation was found for translation in predicted brain areas of executive control e.g. caudate nucleus.”

Similarly, a study called “A Functional Imaging Study of Translation and Language Switching” by Cathy J. Price et al. found that “Translation, but not switching, increased activity in the anterior cingulate and subcortical structures whilst decreasing activation in several other temporal and parietal language areas associated with the meaning of words.”

Although this is by no means an exhaustive list, these fMRI studies agree that the anterior cingulate and specific subcortical structures (caudate nucleus and putamen) are activated during translation but not during monolinguistic tasks. What is interesting is that these areas are specifically devoted not to language but to functions such as attention control, working memory, planning, reasoning, cognitive flexibility, problem-solving, etc. — collectively referred to as executive control.

An analogy with mathematics might be helpful here. Einstein always maintained that his mathematical theories did not come to him in words but as non-verbal concepts and images. fMRI experiments have since confirmed this phenomenon. Researchers Stanislas Dehaene and Marie Almaric found that “in professional mathematicians, a specific set of areas was activated when they were thinking about math problems. […] These areas weren’t activated by non-math problems, and were not activated in the non-mathematicians who treated complex mathematical statements as gibberish.” (“Origins of the Brain Networks for Advanced Mathematics in Expert Mathematicians”). It turns out that basic arithmetic facts and calculations are stored in the left parietal lobe, while complex mathematical reasoning activates the right parietal lobe.

In the same way, linguistic facts and rules are not stored in the same areas that are activated for more complex translation processes. Ergo, reading comprehension and vocabulary tests are not predictors for translation ability.

I think that the elusive “X-factor” is a highly developed executive-function area (in particular the anterior cingulate, putamen and caudate nucleus). And that is something you can actually test. In fact, there are scores of psychoneurological tests that focus on specific brain activities, including executive functioning. My guess is that certain types of executive function tests will yield different results for translators, bilingual non-translators and monolingual non-translators.

To test this idea, I set up a short online survey with three standard cognitive assessment games that are associated with different parts of the brain. The whole thing takes about 10 minutes, and my hope is to get enough participants from all three categories to tell me whether I’m on to something or completely off-base.

Please click here or on the link below to take the survey, and kindly forward this to anyone else you think might be interested! One thing: I apologize in advance for the chipmunk voice in the second game; that’s not me being weird, it’s the sound effect that came with the game!

Once I collect enough data I will be back with the results.

Thank you and have fun!

Go to Translation Brain survey