by Althea Kaminske

(Cover image by Gordon Johnson from Pixabay)

This post continues from a post I did last week discussing whether there is a difference between memory and learning and a difference in learning in the Arts and Sciences. You can find the first part of the post here.

Imposing order in a chaotic world: Subjective organization and running times

Let’s turn back to the idea that there’s a difference between learning and memory. Recall the experiment I described last week about the child's dinosaur knowledge (1). Did the child first memorize a list of dinosaurs and then move on to the next step of learning about these dinosaurs? It’s possible, the researchers did not outline how he acquired his knowledge. My intuition, however, is the 4 ½ year old’s learning process wasn’t as linear as that. Instead, I would imagine that he first learned a few isolated, item-specific, facts about dinosaurs. Then, through talking about them or reading about them, he made some connections, relations, between the dinosaurs which provided a framework to help him remember more about the dinosaurs. It is difficult to obtain the knowledge framework, the relational processing, that helps memory without engaging with the material in some way. I suspect that he was able to build an impressive semantic network of dinosaurs through his constant retrieval practice in the form of talking about dinosaurs with his parents all the time.

In this blog we’ve talked a lot about retrieval practice, and mostly about very structured retrieval practice in the form of tests, quizzes, and games. These are forms of retrieval practice that are often studied using cued-recall in the lab. In a cued-recall task a participant is asked to study something, usually a word list, and then given cues, or hints, to test their memory. A cue can be another item on the list, the type of list (for example, “farm animals” could be given as a cue to help remember most of the items on List 1 above), or a question as is often the case on quizzes or tests (e.g., “What literary devices did the author use in this work?”). Another way retrieval practice is studied is to use a free-recall task. In a free-recall task a participant is asked to study a list or a passage and then simply asked to write down everything they can remember in no particular order. In this type of recall task the experimenter does not give cues, or at any rate not very specific ones. But that does not mean that memory is working without a cue. Instead, in free-recall tasks the participants are relying on self-generated cues during the recall task in order to retrieve the information.

How do we know people use self-generated cues during free-recall tasks? Studying the relationship between retrieval and cues is certainly more challenging when the cues are not provided. One way this has been examined in the laboratory as been through what Endel Tulving termed “subjective organization” of free recall (2). He had people read a list of unrelated words and then recall the words in any order (free recall). Participants were presented with the list of words then asked to recall them for a total of 16 times. Each time the words were presented they were presented in a randomized order. Throughout the 16 times participants recalled the words they tended to keep certain groupings of words together, and those groupings became more fixed each time they recalled the list. Each person’s grouping of words was different, subjective, but they all tended to get more fixed the more times they recalled it. They were imposing structure on an otherwise unstructured list. Tulving called this subjective organization and theorized that they were building associations between words that were grouped together, allowing the words within the list to become cues for other words within the list. They were developing relational processing within the list of words to help them memorize the words.

In the experiment described above, participants were memorizing a list of words. It’s easy to imagine that they were able impose relational processing by coming up with some sort of small story or meaningful association between words. They were given words like finding, treason, office, etc. that might be easy to make some sort of connection between with enough repetition. But what if you’re being asked to memorize something that you can’t make up a story about, like numbers? Even when dealing with memorizing numbers, people will tend to impose structure and meaning on the numbers. People can typically remember a string of about 7 numbers without having to impose some sort of structure or mnemonic device (3). Think about being told a telephone number and then writing it down within a minute or two – you might repeat it to yourself a few times while you search for a pen and paper, but you can hold about 7 numbers in your mind at a time. Researchers refer to the amount of numbers that you can successfully remember over the short term as your digit span. Again, most people can remember about 7 plus or minus two. However, one study reported on someone who could remember 80 digits (4).

How was he able to remember so many numbers? Was he simply better at rote memory than other people? The researchers, Ericsson and Chase, explain that he had a strategy that helped him associate the numbers in a meaningful way (4). The participant was a long distance runner, so he thought of the digits as running times in order to give tem more meaning and group them together. Even in tasks that seem to rely purely on rote memory, we find that people build associations, context, and relational processing to help them remember the information.

Earlier, I argued that memory and learning are an iterative process. Now that I’ve reviewed some of the research on item specific and relational processing I want to explain that process again. When people distinguish between memory and learning they are often expressing a difference between memorizing facts (rote memory) and a deeper understanding of relationships, theories, and processes behind those facts (learning). In this view learning can often be seen as a linear process where students first memorize basic facts, like definitions of words, that they then use later when they learn about the theories or processes in that field. Furthermore, exceptional learning is felt to have taken place if students can generate new ideas or theories based on that learning. This postulates a one-way street from learning facts to learning theories to generating ideas. This is an incomplete picture of learning.

Learning does not happen linearly. As students learn more about an area, their memory and understanding for basic facts is strengthened through the relational processing they’ve gained through exposure to theories and processes. This, in turn, allows them to learn more “basics”, which expands their understanding of theories and processes. Each time a student revisits a concept, their intervening experiences change the structure and framework of their knowledge, allowing for a different understanding of the concept. Hopefully, if their learning has been structured well, this restructuring of knowledge has happened in a way that improves their understanding of the concept.

Since these are very abstract concepts, I made the figures below to hopefully explain this process a bit more. The green nodes represent ideas or concepts. Darker nodes indicate a better memory for that idea or concept. The purple lines represent associations or relational processing between the ideas and concepts. Darker and thicker lines indicate a better understanding of the relationships between and among the ideas.

When people tell me that they don’t want their students to just memorize unrelated facts by practicing retrieval, I try to explain that not only do I not want that either, but also I don’t really know how to do that anyways. To improve memory is to improve learning, and to improve learning is to improve memory. I don’t know how to make those green nodes darker without changing the purple lines.

Now, it’s entirely possible that students can seemingly memorize information without sharing the same kind of deep knowledge of the material that their instructors hope they have. Students can use all manner of mnemonic devices, “memory hacks”, and subjective organization to get the job done. However, in my opinion, these cheap tricks are poor substitutes for the learning and memory that we hope to have in our classrooms. Not only are they less effective in the long run, but they are more difficult to achieve over the short term. Effective memory practices – retrieval, dual coding, concrete examples etc. – are effective because they capitalize on the active process of memory and forming associations between cues and targets. Effective instruction, then, takes these principles into account and helps guide students in learning these concepts and associations through repeated retrieval, elaboration, and interleaving.

Art vs Science

The other part of this argument is that learning is somehow a different process in the Arts than in Science. While I will argue strongly that learning is the same process in both Arts and Science it’s certainly true that there are differences between the Arts and Sciences. Disciplines that focus on testing hypotheses and generating accurate predictions and theories are categorized Science. Disciplines that focus on analysis and expression are categorized as Arts. Despite these differences in goals and process, both are enterprises taken on by humans who learn in fundamentally the same way. As such, the ways in which we individually pursue and succeed at Arts and Science end up being very similar, despite our misconceptions about both fields.

Disciplines in the Arts are often portrayed and seen as being very subjective, somewhat unstable, and less disciplined and more dependent on innate talents or gifts than those in the Sciences. People often talk about genius in artists, writers, and performers as something that works through them. Something that is often outside of their control and is gifted to them.

Disciplines in the Sciences are often portrayed and seen as being very objective, uncreative, and the more dependent on hard work and diligence than those in the Arts. People often talk about the solitary genius of scientists who persist at solving puzzles and dreaming up solutions to the universe’s greatest mysteries.

Neither of these views is accurate or representative of these fields. A talented or gifted artist doesn’t suddenly produce a great work out of nowhere. Many unseen hours are put into practicing and understanding their craft. Art is not entirely subjective and random, it is not without rules and principles. In the case of literary works these principles and rules come from an understanding of language, society, and culture. In the case of visual arts these rules and principles come from an understanding of the media being used as well as how the human eye experiences a piece. In the case of music these rules and principles come from an understanding of composition, instruments, and how the ear experiences a piece. To be a great artist is to control all those elements to subvert or satisfy expectations.

A great scientist doesn’t solve the mysteries of the universe by working steadily alone, patiently following rules and procedures. Science is not entirely objective and uncreative, it is much more collaborative, creative, and subjective than most are willing to admit. Science, at its heart, is a collaborative effort. Scientists share and discuss ideas at conferences, in classrooms, online, or any other forum they can get their hands on. New processes and methodologies are created to answer new questions generated by these discussions. The validity of the questions and the answers are determined by members of the scientific community and, increasingly, the public. To be a great scientist is to create and communicate ideas clearly.

Both the Arts and the Science require individuals to learn the various methodologies and theories within their disciplines. Great artists and great scientists develop skills and deeper understanding through practice and feedback. Experienced artists and scientists organize their knowledge differently than inexperienced artists and scientists and their understanding of basic concepts changes and deepens as they continue on in the discipline. Both require memory and learning.

References:

(1) Chi, M & Koeske, R. D. (1983). Network representation of child’s dinosaur knowledge, Developmental Psychology, 19(1), 29-39.

(2) Tulving, E (1962). Subjective organization in free recall of “unrelated” words. Psychological Review, 69(4), 344-354.

(3) Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 101(2), 343-352.

(4) Ericsson, K. A. & Chase, W. G. (1982). Exceptional memory. American Scientist, 70(6), 607-615.