A Dutch study proposes that the brain is a prediction machine that tries to anticipate what is happening in our environment.
Our brain works a bit like the autocomplete function on the phone or Google: it constantly tries to guess the next word. This happens every time we listen to a podcast, read a book, or chat.
Our brain is constantly making predictions at different levels, from meaning and grammar to specific speech sounds. This is what researchers at the Max Planck Institute for Psycholinguistics and the Donders Institute at Radboud University in the Netherlands managed to establish. The conclusions of Their study was published in the journal PNAS.
This is consistent with a recent theory about how our brains work: what we have between our ears is a prediction machine, continually comparing the sensory information we receive (such as images, sounds, and language) with internal predictions.
According to the study’s lead author, Micha Heilbron, “This theoretical idea is very popular in neuroscience, but existing evidence for it is generally indirect and restricted to artificial situations.” Brain research on this phenomenon is usually done in an artificial environment, says Heilbron.
To evoke predictions, participants are asked to look at a single pattern of moving dots for half an hour, or listen to simple patterns in sounds such as “beep beep beep, beep beep beep, ….”. There are several studies that reveal that our brain can make these predictions, but until now it had not been verified whether this applied to the complexity of everyday life situations, which are not simple artificial series of sounds or images.
To prove this, researchers analyzed the brain activity of people who had heard Hemingway or Sherlock Holmes stories. At the same time, they analyzed the texts of the books using computer models. In this way, they were able to calculate for each word in the text how unpredictable it was.
For every word or sound, the brain creates detailed statistical expectations and turns out to be extremely sensitive to the degree of unpredictability: the brain’s response is strongest when a word is unexpected in context. This is not surprising, the brain sometimes “fills in” the blank and mentally completes another person’s sentences when, for example, he speaks too slowly, stutters, or can’t find the right word.
The experiment was able to verify that, in fact, we do this continuously, under any circumstances. In this sense, the brain performs a task similar to that of the dictation interpreter or the phone’s auto-completion function, trying to know what comes next according to the expectations generated by what it has heard or seen before.
However, there is one big difference: brains don’t just predict words, they make predictions on many different levels, from abstract meaning and grammar to specific sounds in a song.
REFERENCE
A hierarchy of linguistic predictions during natural language understanding