WHAT CAN A CHESS GRANDMASTER teach us about memory?
Chess has been popular research topic which has allowed theorists to study the limits of human memory capacity by comparing chess grand masters’ ability to store and retrieve information with novice chess players.
[@more@]
How do they do it?
Chess grand masters have demonstrated their extraordinary abilities by simultaneously playing hundreds of boards and winning them all. One grand master said, “I see only one move ahead, but it is always the correct one.”
Chess grand masters rely more on their structured knowledge than on extended analysis. It’s the first few seconds of thought that make the different between a novice and a grand master (Ericcson). This rapid, knowledge guided perception, apperception, is a model for studying the limits of memory (Simon). How do chess grand masters overcome the limits of memory to achieve their vast storehouses of information?
Meaningful chunks
Miller (1956) described short-term memory as a small waiting room leading into long-term memory; only about seven items could be in a short-term room at a time. If one more item was attempted, it wouldn’t find any room or previous items would have to fall out or move into long term memory. Miller described how chess grand masters can overcome the magic number 7 by proposisng that the information need not be single items but rather could be “chunks.”
Chase and Simon (1973) extended Miller’s theory and said that information was stored in the brain in chunks. They also said chunks can be arranged in highly organized, intricately structured and complex systems of meaningful patterns and templates and could be recalled as a single unit. Novices have collections of individual pieces stored in memory while grand masters have collections of familiar meta-configurations. Because expert memory is an indexed collection of these units, chunks are treated as single items in long term memory and can be stored and retrieved in one action.
Cognitive load for the expert and the novice
A chess grand master can recall a complex board set-up after only a few seconds of examination while novices could only recall a few details after 30 seconds of exposure to the same set-up because grand masters can search more deeply along the branching trees of possible moves (Gobbet and Clarkson, 2004). A novice sees 20 chess pieces and thousands of possible chunks. A grand master sees the same board within the context of larger patterns of information: the entire setup could be perceived in 5 or 6 chunks.
Instructional design and cognitive load
Gladwell in his book Blink (2005), calls it “adaptive unconscious” and says we all have “rapid cognition” to some extent. Expert performers such as chess masters provide deeper insights into the limits of cognitive load and memory. If expert memory systems can be more precisely described, predictions might be refined for the benefit of improved instructional design.
Fink, Diana, (2006). Chunks: Superior memory of experts. In B. Hoffman (Ed.), Encyclopedia of Educational Technology. Retrieved September 24, 2008, from http://edweb.sdsu.edu/eet/articles/chunks/start.htm
西洋棋專家可以教我們什麼有關記憶的事？
西洋棋是熱門的研究主題，允許理論家學習人類記憶容量的限制，藉由比較西洋棋專家和新手的儲存和擷取資訊的能力。
他們如何做的？
西洋棋專家展示他們非凡的能力，同時進行上百個棋局，而且全部戰勝。某個專家說「我只看事前一步，但它總是正確的一步」
比起廣泛的分析，西洋棋專家較信賴他們結構化的知識。新手和專家間的不同是一開始的少數幾秒的想法(Ericcson)。這個迅速的，知識引導感覺、統覺，是學習記憶限制的模型(Simon)。西洋棋專家如何戰勝記憶的限制去達到他們廣大的資訊容量？
有意義的Chunks
Miller(1956) 描述短期記憶像要前往長期記憶的一個小的等待空間。在短期記憶的空間同時只有大約七個項目可以存在。如果多嘗試一個，它將找不到任何空間或前一項可能會減退和移到長期記憶。Miller 描述西洋棋專家如何可以戰勝這奇妙的數字7，藉著提出資訊不需要單一項目，而可以是複數的「chunks」。
Chase 和 Simon (1973) 擴展 Miller 的理論，並且認為資訊以 chunks 的方式儲存在大腦裡。他們也認為 chunks 可以安排較高的組織，複雜的結構和複雜的有意義樣式的系統與樣本，並且可以像單一單位般的回憶。新手聚集儲存在記憶中個別的部份，當專家是聚集相似的後設結構。因為專家的記憶是這些聚集單位的索引，chunks 被視為長期記憶中的單一項目，並且可以在單一活動中儲存及回憶。
專家與新手的認知負荷
一個西洋棋專家可以回憶一個複雜的棋盤方案，只要少數幾秒的調查，當新手在同樣的棋盤方案要花30秒，只能回憶少數細節，因為專家可以延著可能的分支樹中做較深的搜尋(Gobbet and Clarkson,2004)。新手看20個西洋棋的部份，和許多個可能的 chunks。專家較大的知識形態看相同的棋局脈絡，其完全的組織可以察覺5個或6個 chunks。
教育上的設計和認知負荷
Gladwell 在它的書 Blink(2005)中，稱為「適應潛意識」，並且說我們在某些範圍都有「快速認知」。熟練的實行者像西洋棋專家提供較深的理解在認知負荷與記憶的限制。如果熟練的記憶系統可以較準確的描述，對改善教育設計好處的預測可能是精確的。