Behavioral and brain responses to language reflect different levels of linguistic representation
This study demonstrates that while both behavioral and neural language responses are significantly predicted by processing effort, only neural data uniquely capture rich, high-dimensional linguistic content beyond effort, suggesting that brain activity reflects detailed comprehension dynamics that behavioral measures bottleneck into simpler properties.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Human language is a constant stream of sound and meaning, but for scientists trying to understand how the mind handles it, there are two main ways to look at the process. One way is to watch what people do: how fast they read a sentence, or how long they pause before answering a question. This is behavior, the visible output of thinking. The other way is to look inside the brain itself, using machines that map electrical signals or blood flow to see which parts light up when a word is processed. For a long time, researchers have wondered if these two windows into the mind show the same picture. A popular idea suggests that both our actions and our brain activity are driven mostly by how much mental effort a word requires. If a word is rare, very long, or completely unexpected in a sentence, it takes more work to process, and both our reaction times and our brain signals should reflect that extra strain. This view treats language processing as a matter of efficiency, where the brain and the body simply react to the difficulty of the task at hand.
However, another perspective argues that the brain is doing something much richer than just calculating effort. It suggests that neural activity captures the deep, complex meaning of words and how they fit together in a specific context, far beyond just how hard they are to figure out. To settle this question, a team of researchers decided to test these two ideas side by side using a massive collection of data. They gathered results from eight different studies on human behavior and five different studies on brain activity, including four that used magnetic resonance imaging to see the whole brain and one that measured rapid electrical changes on the scalp. Instead of guessing which factors mattered, they used powerful computer programs trained on vast amounts of text to act as a measuring stick. These programs could generate two different types of numbers for every word in a sentence: one set that estimated how much effort it would take to process the word, and another set that captured the full, nuanced meaning of the word within that specific context.
The researchers then fed these numbers into their analysis to see which set of numbers better predicted what actually happened in the human studies. They found that the effort-based numbers were indeed very good at explaining the data. Whether looking at how fast people read or how their brains reacted, the difficulty of the word accounted for a large portion of the results, confirming that mental effort is a major driver of language processing. But the story changed when they looked at the brain data more closely. While the effort numbers explained the brain's reaction well, the rich, meaning-based numbers added a significant amount of extra explanation. In other words, the brain was responding to the deep meaning of the words in a way that the simple measure of effort could not capture. This extra layer of detail was not found in the behavioral data. When the researchers looked at how people acted or reacted, the effort-based numbers were enough to explain the results, and the rich meaning numbers did not add any new predictive power.
This difference suggests that the brain and the body are telling two slightly different stories about the same event. The brain appears to hold a high-resolution, detailed record of language, tracking the complex web of meaning and context as it unfolds. The behavior we see, however, seems to be a simplified version of that process. It is as if the mind takes all that rich, complex information and funnels it down into a few key properties before it reaches our actions. The brain sees the full landscape of meaning, but our behavior only reflects the steepness of the hills we have to climb. These findings suggest that while both brain and behavior are influenced by how hard a word is to process, only the brain provides a direct window into the intricate, high-dimensional dynamics of understanding language. The body gives us a useful summary, but the mind holds the full picture.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.