‘How is language processed in the brain by native speakers of different languages? Is there one brain system for all languages or are different languages subserved by different brain systems? The first view emphasizes commonality, whereas the second emphasizes specificity’.
This is the opening statement of a research article published in the January, 2015 issue of PNAS in which scientists have investigated how our brain processes two very diverse languages: a tonal language (Chinese) and a non-tonal language (English). A group of researchers from China and the UK, led by Jianqiao Ge from Peking University, Beijing, have performed a study on 30 native Mandarin Chinese speakers and 26 native English speakers.
In English and non-tonal languages in general, pitch modulation is used to express emotional information. In contrast, pitch modulation in Mandarin and other tonal languages indicates different words altogether with distinct meanings. For example, the word Bi in Mandarin could mean force, nose, wall or compare, depending on how you pronounce it. Processing Mandarin speech therefore requires a higher degree of mapping tone to lexical meaning as compared to English. The researchers claim these differences between Mandarin Chinese and English change the way the brain’s networks work.
The classic brain regions associated with lexical speech processing are the Broca’s and Wernicke’s regions found in the left cerebral hemisphere. The corresponding regions in the right hemisphere have been implicated in the emotional speech processing. These findings were made, before real time functional brain imaging such as functional magnetic resonance imaging (fMRI) was possible. Brain autopsies of patients with speech impediments revealed lesions to these regions.
Previously, using fMRI to study the activity of the interconnected brain regions across languages was limited by the huge computational requirements. In this study, Ge and his team used cloud computing to analyse thousands of dynamic causal models to map the flow of activity between the regions involved.
The experimental task exposed the subjects to previously audio recorded intelligible and unintelligible phrases from their native languages. The phrases were voiced by a male and a female who are native speakers. The subjects were required to identify the gender of the speakers.
The results reinforced previous findings in general speech processing, showing activity in three regions of the left hemisphere namely, inferior frontal gyrus (IFG/Broca’s area), the anterior and the posterior temporal gyri (aSTG and pSTG/Wernicke’s area). The sound signals for both the groups entered the neural system via the pSTG.
However, there are some differences in the nature of the interactivity between the regions. In English speech processing, the pSTG to IFG connection is stronger, whereas the aSTG connections to both IFG and pSTG are stronger in Mandarin speech processing.
Another significant difference is the recruitment of the right hemispherical aSTG by Mandarin speakers, but not by English speakers. The right aSTG has active connections to both the classic left regions in Mandarin speakers.
These findings emphasize the importance of developing a bilateral network between the two brain hemispheres to speak and understand languages, particularly for tonal languages like Mandarin Chinese.
Extending this research to cover another feature called “pitch accent”, associated with certain languages, could increase the comprehensiveness of understanding speech processing. Scandinavian languages and Japanese are examples of pitch accent languages, where stressing on syllables alters the meaning of the word. However, such pitch alterations are restricted to one or two syllables in the word. This differs from fully tonal languages where each syllable has its own tone.
There are more than 7000 different spoken languages currently in use. With advancements in computational capabilities, the time is ripe for the neurolinguistics field to expand cross-language studies to include several languages.
The original paper can be accessed here.