Welcome to Memory Mondays, where I read a textbook on memory and talk about what I Iearned. If you like your cognitive psychology neatly summarized, with a healthy dose of unnecessary commentary and excessive amounts of semi-colons, this is the series for you!
In the previous Memory Mondays post, I talked about the value of consolidation and semanticization of memory, especially in school. While about episodic memory, this chapter still holds a lot of information relevant to semantic memory and school learning, to the point where I don’t really understand why the title of the chapter used ‘episodic memory’ at all.
Episodic memory is memory of moments that happened at a specific point in the past, and has been likened to mental time travel, in that we can relive earlier events and anticipate future events. For episodic memory (or any explicit memory, I think) to function properly, it needs a way to encode that experience so it is distinct from others, a way of storing the memory, and a way to search for and retrieve it.
Ebbinghaus and Bartlett
There are two broad traditions in the study of memory. At least, there was in the past; I don’t think more current research can be neatly split into two schools.
If you are a frequent reader of this blog (and thus a member of a very small group), then you may already be familiar with Ebbinghaus. He was an early pioneer in cognitive psychology, demonstrating that the mind, specifically memory, can be studied in the laboratory. The main criticisms of his research is that the material he used is inauthentic, strings of nonsense syllables, and that the outcomes of his studies tell us little about how memory works outside the laboratory.
Bartlett was one of these critics. He argued that the meaning of the material matters, and conducted his research using complex material such as stories from different cultures. Bartlett was mainly interested in the impact of schema: how long-term structures of knowledge in our minds can influence how we understand new information and thus how we store and retrieve it. Participants in his studies would read and later recall stories from another culture; the errors they made typically adjusted the stories to more closely align with their own cultural expectations of a narrative. The occurrence of schema-driven errors is greater the longer the interval between encoding and retrieval, as schematic information lasts longer in memory, while novel information will fade. Subsequent research has also shown that the influence of schema is larger with recall than with recognition tasks.
Schema do not only create errors in memory; they also serve a useful purpose in increasing the accuracy of memory, if the new material can somehow be linked to what is already in the schema. Participants in one study were asked to recall (and draw) meaningless patterns from memory; recall was greatly improved if the patterns when initially accompanied by a label. What was meaningless became meaningful by linking it with schema, and recall improved.
Bartlett criticized Ebbinghaus for removing meaning from the material to be memorized, and thus decreasing the relevance of his research to the real world. However, as later studies have shown, even when memorizing nonsense syllables, it is not possible to entirely remove meaning. Syllables such as CAS can be linked to words like castle, or syllables that more closely follow the structures of English are easier to memorize. For studies using lists of words and free recall tasks, where participants can recall the words in whatever order they choose, words that are more closely associated with each other are recalled together, even if they were separated on the list initially learned. As well, if the words evoked an image, such as canoe, they are easier to remember than more abstract words like virtue. Meaning influences memory, even when researchers think it to be more secondary to the task.
The implications for teaching are apparent. If we want students to recall the information we teach them, we should endeavour to link it to what they already know. This could be what they learned last lesson, last week, last unit, or last year; the further you stretch back, the better knowledge you need of the school’s curriculum in different grade levels. Students will also benefit from a vertically- and horizontally-aligned curriculum, so as many links can be made as possible (see the 3-D curriculum described here).
Retrieval and Encoding
If we want memories to be accessible and useful, we need for the memory to be strongly encoded.
How the information is initially processed influences the durability of the memory. Craik and Lockhart promoted this idea with their levels of processing theory. They got participants to process words visually (how they looked), phonologically (how they sounded) or semantically (by their meaning), with the semantic processing assuming to have the greatest depth. Those who processed the words semantically had the better memory for the words. This effect holds even if processing time is equal for all the tasks and regardless of whether participants expect a test. The main criticism of the theory is that the different features of a stimulus may be processed simultaneously, rather than sequentially in increasing depth; that is, we take in the sight, sound, and meaning of a word at the same time.
Deeper coding can also be achieved through elaborative rehearsal, such as by linking the new material to other material held in memory. Organizing the material is also helpful, such as in a hierarchy, into categories, or associating it with visual imagery or a story.
As more information is deeply encoded in the mind, the more elaborate schemas grow and the more useful they become for our thinking processes. The most interesting part of this chapter was about long-term working memory, where structures of knowledge develop in the long-term memory, where they can be more easily accessed and used within working memory. This is the magic of expertise: easy access to complex knowledge and skills, which can then be utilized in critical thinking, creative, and problem-solving processes.
So, as you can see, despite the label of episodic memory, there was a lot in the chapter applicable for building semantic memory and schemas. And there will be more of that goodness…
Next week: knowledge and semantic memory!