The melodic brain: How we stay tuned to vocal pitch in speech

In everyday conversations, we are not only paying attention to what the words mean but also how the words are expressed. For example, when a colleague just dropped by our office and said “Coffee?” with a vocal pitch rise, we probably would know in an eyeblink that this is an invitation to have some coffee together – more than simple information about a hot tasty drink in the kitchen.

by Pei-Ju Chien

Most languages use pitch cues in the form of speech intonation to make a statement or a question. Interestingly, more than a half of the world’s languages are so-called tonal languages, such as Mandarin Chinese, which also use pitch cues to convey word meaning. This specific function of vocal pitch for word meaning is termed lexical tone. A Mandarin word may refer to completely different things when the lexical tone differs. For instance, /ma/ can be ‘mother’ when spoken with a Mandarin Tone 1 (i.e., high flat pitch contour), but then becomes ‘a horse’ with a Tone 3 (i.e., low falling-rising pitch contour).

As a tonal language is “twice melodic” with tone and intonation, it is now getting complicated to imagine how a native speaker decodes respective information from overlapping acoustic cues: One needs to grasp the vocal pitch for the word meaning itself (‘coffee’) AND what is said between the lines (‘a coffee break?’). How does the brain make sense of that? From previous studies we know a bit about how this works in speakers of non-tonal languages, but how does it look for tonal language speakers?

Our recent study (Chien et al., 2021) aimed to answer these questions by investigating 24 Mandarin speakers with functional magnetic resonance imaging (fMRI). Our results showed that Mandarin speakers’ brains have similar activation in auditory regions keeping track of tone and intonation. But, they are particularly engaged in strong inter-regional communication to support precise intonation processing.  

When participants were instructed to identify tone (Tone 2 or Tone 4) or intonation categories (statement or question), increased brain activity in bilateral auditory regions was commonly observed during both tasks. This finding may indicate that tone and intonation are processed similarly to meaningful pitch-related categories in Mandarin. Importantly, beyond these similarities with tone, intonation involves additional activity in the frontal regions and stronger processing integration between activity in the frontal and auditory areas. Increased processing integration was also found between activity in the frontal and motor areas. These results may show that intonation is represented in the brain differently than tone, having extra support for evaluating pitch contours and preparing a verbal response. 

Together, our fMRI study demonstrated how the brain works with the dual function of pitch cues in a tonal language. Native speakers are capable of staying tuned to vocal pitch by relying on the auditory cortices for category representation, irrespective of its relevance to word meaning (tone) or communicative messages (intonation). Yet, specifically for tuning to intonation, tonal language speakers need additional recruitment of fronto-auditory and fronto-motor inter-regional communication for further pitch contour labelling and vocal response planning. What is also interesting is that the overall bilateral representation of Mandarin intonation is similar to previous findings on intonation in non-tonal languages, such as English or German. An important next step is to understand whether processing vocal pitch for intonation is generalizable, by comparing speakers with different language backgrounds. Our previous study (Chien et al., 2020) suggests that German speakers may actually process Mandarin intonation in a very similar way as Mandarin speakers. Although further cross-language research is clearly necessary, our results hint at a possible scenario. Perhaps, no matter which language it is, our brain simply stays tuned and is ready to have some hot and tasty drinks together!

Veröffentlicht von

Pei-Ju Chien hat Englisch und Linguistik in Taiwan studiert. Derzeit promoviert sie am Max-Planck-Institut für Kognitions- und Neurowissenschaften in Leipzig. Als Muttersprachlerin einer tonalen Sprache, bei der mit einer Änderung im Ton auch eine Änderung der Bedeutung des Wortes einhergeht, interessiert sich Pei-Ju besonders dafür, wie unser Gehirn "Sprachmelodien" in alltäglichen Gesprächen wahrnimmt. In ihrem Promotionsprojekt untersucht sie mithilfe der funktionellen Magnetresonanztomographie (fMRI), welche Hirnregionen bei tonalen Sprachen beteiligt sind. Außerdem will sie sprachübergreifend herausfinden, ob und wie sich die Hirnstrukturen zwischen Sprechern tonaler und nicht-tonaler Sprachen, zu denen auch Deutsch gehört, unterscheiden.

2 Kommentare

  1. Quotes from above:

    In everyday conversations, we are not only paying attention to what the words mean but also how the words are expressed.
    ….
    What is also interesting is that the overall bilateral representation of Mandarin intonation is similar to previous findings on intonation in non-tonal languages, such as English or German.

    My Interpretation: In everyday conversation, we are not only paying attention to what the words mean but also what the overall message is.
    In short: we are paying attention to the communication intention of the speaker. Because of that, brain activation in a mandarin hearer and a german hearer will be in the end quite similar.

    • Thank you for your feedback, that’s an interesting idea! Indeed, there have been some accounts addressing the role of communicative intention recognition in language comprehension (e.g., Catani & Bambini, 2014; Lima et al., 2016). Specifically, previous literature investigating communicative intention recognition via prosody suggests that the relevant brain networks involve auditory prosodic, socio-cognitive and cingulo-opercular brain areas (e.g., Hellbernd & Sammler, 2018). Further empirical evidence (including cross-linguistic comparisons) is definitely desired to support the current understanding of intention recognition.

Schreibe einen Kommentar


E-Mail-Benachrichtigung bei weiteren Kommentaren.
-- Auch möglich: Abo ohne Kommentar. +