Things to Remember:Limits, Codes, and the Development of Object Working Memory in the First Year
Alan M. Leslie, Zsuzsa Ka.Loy · 2007
Abstract The cognitive revolution of the 1960s had its roots in breakthroughs in the mathematics of machine theory, cybernetics, information, and formal systems associated with Turing, Weiner, van Neumann, Shannon, and Chomsky. For the first time, it was possible through the science of computation to glimpseat least in principle-a physical basis for mind and mental phenomena. When Miller (1956) published his famous article on memory, it was to proclaim two things that have now become familiar. The first was that we can actually measure mental phenomena without white coats and large machines but, best of all, without embarrassment. The second was that some human memory systems have an astonishingly severe limit. There was also a third finding. Though we can (fairly) easily attach a number to this limit, and measure it in that sense, it is much harder to characterize what exactly are the twits that are the subject of this limit. Miller believed the number was around seven; but he quickly showed that it was not seven of Shannon’s bits but seven of something else, which he called chunks. The human memory system was not simply a channel carrying input passively but a complex system that actively recoded its throughput, as digits, letters, words, phrases, and so on, according to the goals of the task. The limit of seven reflected the structure of these codes. The twin properties of limit and code still define short-term memory research: how much can be stored and how much of what can be stored.