This is a quick note, before I forget the last couple things that I read.
Where did working memory come from? Here’s my picture so far:
- There’s a limit on how much random stuff you can remember at once. You don’t need science to know this; you just need random stuff that you have to remember. I assume this has been known forever.
- People had different names for this. William James called it primary vs. secondary memory. Others were calling it long-term vs. short-term store. What was controversial was whether these constituted just two facets of the same memory system, or whether they were two totally independent memory systems.
- Evidence for independence comes largely from patients with brain damage. These patients either are amnesic, but do perfectly fine with short-term memories, or else they have greatly impaired short-term memories but otherwise function and learn OK. This suggests independence.
- Question: does short-term memory constitute a working memory for functioning and long-term memory? In other words, is short-term memory necessary for learning, reasoning and comprehension?
- Baddeley and Hitch do a battery of experiments to show that impairing short-term memory with verbal info does (modestly) impair learning, reasoning and comprehension. This is evidence that short-term memory does constitute a working memory system.
- But the thing is that performance was only modestly impaired by their experiments, so there must be more to the cognitive system than what their experiments uncovered. (Their experiments almost entirely used span tasks that ask people to remember a bunch of stuff, random numbers, letters, etc.)
- They then go beyond their experiments to make a guess about what the structure of working memory might be. They propose that there is a capacity to just passively take in verbal info, up to a point. Beyond that point, a “central executive” has to take active steps to hold on to info. Thus working memory limitations come both because the passive store has a limited capacity and also because the central executive can only do so much at once. The span experiments fill up the passive store, and force the active executive to do something. This allows task performance to go OK, but at a cost in performance. They also guess that there is a spacio-visual store that is entirely parallel to the phonological store.
- This takes us up to, like, 1974 or something.