When we read a sentence, we need to grasp the structure of the sentence while understanding the meaning of each word. Two different networks are responsible for this in our brain, a study shows. One is activated more and more in the course of a sentence, but not with incoherent lists of words. The other, on the other hand, reacts particularly to individual words. Both networks are in close contact and thus enable us to read fluently.
The language network in our brain ensures that we can understand and produce language, both spoken and written. It is already known from previous studies that the network extends across the frontal and temporal cortex of the brain. Individual regions each have their own functions, but are closely linked. However, how exactly the different regions cooperate in understanding written language has been unclear until now.
Subjects with electrodes in the brain
A team led by Oscar Woolnough from the University of Texas at Houston has now gained new insights into the processes that take place in our brain while we read. To do this, the research team analyzed the brain activity of 36 people as they read different types of sentences or combinations of words: normal sentences, called jabberwocky sentences, which are grammatically correct but made up of made-up words, and word lists of either real or nonsense words.
All subjects were epilepsy patients who had had electrodes surgically inserted into their brains. This enabled Woolnough and his team to record brain activity much more precisely than would be possible with scalp recordings. “Implanted electrodes in the brain give us an unparalleled insight into the inner workings of the human mind, especially during fast processes like reading,” explains Woolnough’s colleague Nitin Tandon. Woolnough adds: “This study helps us to better understand how distributed nodes in the brain’s language network work together and interact to enable us to understand complex sentences. In order for us to be able to understand language, a precise sequence of fast, dynamic processes must take place in different parts of our brain.”
Two interacting networks
In fact, the subjects’ brains showed different activation patterns depending on whether they were reading meaningful sentences, Jabberwocky sentences, or word lists with or without meaning. “We discovered two functionally distinct networks spanning the frontotemporal cortex,” reports the research team. The first network thus includes a region in the frontal lobe of the brain that sends signals to the temporal lobe. “Activity in this network increases over the duration of a sentence and is reduced or absent in Jabberwocky and word lists, suggesting a role in capturing meaning at the sentence level,” the team explains.
The second network includes another region of the temporal lobe that sends signals to an area of the frontal lobe. In contrast to the first network, this network is mainly active when reading word lists, but less so when reading entire sentences. Woolnough and his team conclude that this network is primarily responsible for capturing the meaning of individual words. When the words are embedded in a sentence, the context makes processing easier, so there is less demand on this second network.
Dynamic billing
“These juxtaposed but spatially and temporally separated neural mechanisms for word- and sentence-level processing shed light on the multi-layered semantic networks that enable us to read fluently,” the team writes. “These results suggest a distributed, dynamic computation in the frontotemporal language network rather than a clear dichotomy between the contributions of the frontal and temporal structures.”
Source: Oscar Woolnough (UTHealth Houston, Texas) et al., Proceedings of the National Academy of Sciences, doi: 10.1073/pnas.2300252120