Here's the transcript from the Science podcast on this topic:
Amateur or professional? Novice or expert? Which are you?
According to the National Resource Council, one of the differences between novices and experts is that "experts notice features and meaningful patterns of information that are not noticed by novices."
Quantifying that ability to recognize features and patterns, however, is difficult. But now, in a paper in this week's Science, Keiji Tanaka and colleagues have shown that experts use different parts of their brains than novices do when engaged in their expert subject. Science's Sophia Cai has more.
Feature Writer - Sophia Cai
Just from everyday experience, most of us intuitively know whether a puddle is too big for us to hop over or how hard to brake when riding a bike downhill. But people who practice something so much that they become experts at it develop a kind of intuition that goes way beyond everyday experience - such as finding solutions to very specific problems very quickly, without stopping to think. An example of such expert intuition is found among professional players of a game called shogi.
Interviewee - Keiji Tanaka
Shogi is Japanese Chess - similar to chess, but a little more complicated.
Feature Writer - Sophia Cai
Keiji Tanaka is a neuroscientist at RIKEN in Wako, Japan, who wanted to understand what part of the brain is used in expert intuition. The hope is that understanding the neural circuitry involved in playing shogi may help in understanding how we process other information, make decisions, and achieve goals, such as troubleshooting a complicated engineering system and finding the one part of thousands that is going wrong.
Interviewee - Keiji Tanaka
And this intuition seems to be field-specific and the result of long-term training.
Psychologists say at least 10 years, everyday, extensive training. But there is almost no knowledge about where in the brain it is performed.
Feature Writer - Sophia Cai
So Tanaka and his team used functional magnetic resonance imaging, or fMRI scans, to look inside the brains of shogi players while they were playing.
Interviewee - Keiji Tanaka
Shogi is very popular in Japan, and there are 160 active professional players. So by using the professional players, and also we imaged the amateur players, we found professional players are clearly different, they make different use of the brain in their problem solving.
Feature Writer - Sophia Cai
And just as you might expect, or intuit, Tanaka and his team found the intuition of experts was different from the intuition of amateurs in the brain.
Interviewee - Keiji Tanaka
We found the precuneus, a part of the parietal cortex, was specifically activated by actual shogi patterns, but not by other objects. So we first found that precuneus is important for perception, quick understanding of the shogi board patterns.
Feature Writer - Sophia Cai
What Tanaka means by "shogi board patterns" is situations in the game that come up during play. And when Tanaka says the precuneus is used for quick understanding, Tanaka means really quick: just seconds.
Interviewee - Keiji Tanaka
Eight seconds is short for me, but long enough for professional players. They make logical thinking, and not only the next move, but they searched the next, next move, including opponent move, several steps. And then searched many possibilities and find the best selection.
Feature Writer - Sophia Cai
Now, both professional and amateur players were able to come up with good solutions, and several cortical areas of the brain were similarly activated in both types of players. But in a follow-up experiment, when the players were required to respond even faster - just a second to see the board and two more seconds to choose the best next-move from among four choices - the amateurs usually failed to choose the optimal move.
Interviewee - Keiji Tanaka
But professional players can report the correct answer in 70% of the trials. We found that,in this case, in professional players, during this quick generation, the basal ganglia was activated. So taken together, these two sets of results, we conclude that basal ganglia is specific to the quick generation.
Feature Writer - Sophia Cai
Psychologist and chess expert Fernand Gobet says it's this activation of the basal ganglia - or more specifically, a region in the basal ganglia called the caudate nucleus - that is key. Gobet works at Brunel University and is not associated with the paper.
Interviewee - Fernand Gobet
The precuneus was found in one of the experiments I carried in chess. That's not totally new. So the real contribution of paper is to show that one specific part of the brain, the precuneus, which encodes perceptual information, and this is linked to another part of the brain, the head of caudate nucleus, which encodes information about actions.
Feature Writer - Sophia Cai
And those actions are-for shogi experts-directed towards the goal of winning the game, even if the best next move means sacrificing a player, or, as is allowed in shogi, using the move to take a player that's already been eliminated and put it back on the board. Again, Fernand Gobet:
Interviewee - Fernand Gobet
The idea is that you first recognize a perceptual pattern, in the first part of the brain they identified, which is the precuneus, so this is purely perceptual. Then activity moves to the caudate nucleus, where you identify a move. And basically you have this stimulusresponse, which provides you an automatic access to a solution.
Feature Writer - Sophia Cai
But figuring out how the brains of experts come to do this - and in such a short amount of time - is still an open question. Study author Keiji Tanaka and his team ruled out habits because there are too many patterns that come up in shogi. They ruled out a lookup table approach that goes through every board pattern ever seen, as even shogi-playing super-computers that use that strategy fail to beat professional shogi players. Instead, Fernand Gobet thinks that shogi game experts use something called "chunks" to help categorize board patterns and strategic moves.
Interviewee - Fernand Gobet
Basically, chunks are perceptual patterns that can be used as a unit. So if you are a beginner and look at a shogi board or a chess board, you recognize only one piece at a time and you can only memorize one piece at a time. If you become a better player, then you can learn to recognize groups of pieces, maybe a bishop together with a knight, or a bishop with a king and two pawns, and so on and so forth. When you learn to read, at least in some languages, you first learn letters, then you learn to put the letters together in syllables, and then you learn about words. So that's exactly the same idea.
So you start with simple units, and you learn to put the units together. And the idea for shogi is what experts do is to learn these perceptual units. And not only do they learn these perceptual units, but these units get attached to information that could be useful.
Feature Writer - Sophia Cai
This idea of becoming an expert by learning these "chunks" is not new: the idea has been circulating for decades, ever since psychologists determined in the 1950s that professional and amateur chess players didn't differ in the depth or breadth of search for the best next-move. But while the world-class players would usually find the optimal move in their search, the amateurs typically would not. Again, study author Keiji Tanaka:
Interviewee - Keiji Tanaka
So this indicates that the idea of the best next-move is not conclusion of the search, but the idea came before the search. And psychologists thought that chunks are stored in long-term memory and chunks should be associated with the idea of the best next-move.
But the question is, where this knowledge is associated and the mechanism of that association.
Feature Writer - Sophia Cai
A question that so far, researchers haven't been able to answer. For Science Magazine, I'm Sophia Cai.
Host - Robert Frederick
You can read Tanaka and colleagues' paper, titled, "The Neural Basis of Intuitive Best Next-Move Generation in Board Game Experts." Find it in this week's Science.
Walter Derzko
Comments