Seeing is believing:we need to see something before we can accept that it really happened. Or do we?
Maybe you have had the experience of completely misreading a sentence and then needing to go back.The reason could be that the brain relies on its prior beliefs and expectations, at the expense of thoroughly processing what it actually read.
Apopular theory in psychology and neuroscience is that our brains are prediction machines. We continuously predict what is coming next, andfocus on information that deviates from that prediction. When predictions are incorrect, we learn to suppress the incorrect prediction, such that we can make better predictions in similar situations in the future.
Research led by Dr Joost Rommers suggests thatpredictions can be very strong, such thathaving expected to read a word can be similar toactually having read the word.
The researchers recorded electrical brain activity from participants who read sentences with unexpected endings (“He swept the floor with a smile”, where “broom” would have been expected).A few sentences later, participants read a different sentence containingthe word they had expected to read earlier(“He saw a man with a broom”). This allowed the researchers to investigate what had happened with the incorrect prediction.
The researchers made use of the word repetition effect: this is a reduction in the neural response to repeated words,relative to words presented for the first time.This reduced response isan indication that a word has been read before. If having expected a word is similar tohaving actually read it, then expected-but-not-seen words like “broom” should act like repeated words, and show a similar reduced neural response. This is exactly what was observed.For further detail, the figure shows electrical brain activity in response to words like “broom” in sentences such as “He saw a man with a broom”. The neural response differed depending on what had been shown a few sentences earlier. In the baseline condition, when “broom” had not previously been seen, there was a large response (blue line). The peak around 400 milliseconds (known as the N400)was much reduced when “broom” had been read previously (red). Strikingly, the peak was also reduced when words had merely been expected, as participants had read “He swept the floor with a smile” a few sentences ago.
For further detail, the figure shows electrical brain activity in response to words like “broom” in sentences such as “He saw a man with a broom”. The neural response differed depending on what had been shown a few sentences earlier. In the baseline condition, when “broom” had not previously been seen, there was a large response (blue line). The peak around 400 milliseconds (known as the N400)was much reduced when “broom” had been read previously (red). Strikingly, the peak was also reduced when words had merely been expected, as participants had read “He swept the floor with a smile” a few sentences ago.
It seems therefore that merely having expected a word is similar toactually reading it.So,seeing is believing, but in a different way: sometimes we see what we believe we will see.Research on prediction holds promise for explaining the mechanisms of language comprehension and other aspects of thinking, perception, learning and memory.
Read full article here.
Dr Joost Rommers, School of Psychology
