Learning: It’s complicated and we know next to nothing about it

Any one who has studied how we learn for any length of time will probably agree on a few things:

  • We know a lot about how the brain learns
  • There is far more we don’t know about the how the brain learns than we do know (known unknowns are greater than the known knowns)
  • A lot of the things we think we know will probably turn out to be wrong (this for example)

These are the things I think I know are true about how the brain learns:

  • One definition of learning is the ability to recall memories.
  • Information gets into our brain through our senses – Auditory, Visual, Tactile, Olfactory and Gustatory (there are other ways of naming these)
  • There are two main types of memory, working memory and long term memory.
  • Working memory is sometimes broken down into a model that includes Visual Working Memory, Verbal Working Memory and a processing capability (and there is some evidence that the processing memory is multi-functional – i.e. it can also be used to extend the working memory).
  • Working memory can be though of as a buffer (for the computing geeks out there). It is usually considered that 5-9 objects can be stored in verbal working memory plus maybe 2-4 in visual working memory at any one time (in comparison, long term memory can be considered, for the purposes of an individuals lifetime, to be limitless).
  • Long term memory has two types, Semantic and Episodic.
  • Generally, episodic memory derives directly from sensory input and is involuntary.
  • Generally, semantic memory stores memory traces from working memory. This includes thoughts, processes and schema from the thinking in working memory.
  • The processing of information in working memory triggers storage in long term memory. Then the brain has to build the ability to easily retrieve the stored memory.
  • Episodic knowledge degrade relatively quickly. By paying attention to sensory inputs, they get introduced into working memory and contextualised. They can then pass into long term memory.

This can be visualised as follows (although given the rapidity with which this area of study moves this is likely to have changed):

If we try to go much further than this we head into the territory of known unknowns as regards details where there is a consensus of opinion. Well, I certainly do.

Every brain and sensory system is slightly different. For example, I’m very short sighted. My eyes don’t work as good as others. My hearing is quite sharp. I have a very poor sense of smell. Such brain/organ differences can arise in a number of ways. Some may be genetic, some due to damage occurring sometime between birth and death (it is interesting that many of the great breakthroughs in understanding how the brain works have come from studying individuals who have suffered catastrophic brain damage that ‘broke’ one or other part of their cognitive function). Not all such differences would have a discernible effect that is replicated in many individuals, others will. There has, for example, been some evidence suggesting that dyslexia is connected to problems with rapid auditory discrimination linked to genetic defects.

Some of these differences from the norm are likely to be connected to the organs themselves. Others will inevitably be due to differences in how my brain organises itself. It is likely that because of this I would be better at remembering anything that requires auditory input than visual input. This is not the same as having a learning style but it may affect, for example, how I deal with information that needs to be processed through the visual working memory. I suspect no part of my optic system work at an optimum level, to say the least. Different subject domains require different mixes of inputs. Some meaning can be transmitted entirely verbally, but where it cannot due regard has to be given to the best way of communicating information with careful consideration being given to the differences individual students may have. As a scientific theory the idea of Learning Styles may not hold much water, but that does not mean that content delivery can be homogenous, nor linked solely to aspects of the content elements. Individual needs still have to be catered for. The important thing here is to have a process that ensures that teachers think through all these needs clearly. Some believe that thinking of learning styles over-complicates this. Personally, I think it is actually even more complicated and requires more thought that it is typically given.

We are fairly clear on the idea that the first phase of learning, that is the passing of information into long term memory depends on the processing of information through working memory then on into long term memory. Retrieval of information works the same way, so there would seem to be a limitation here. There are techniques that can be used to help, for example chunking. Recent research has also shown that constant use of that technique can improve the base working memory size (or at least currently give that impression). So perhaps one way to improve learning is to encourage the use of chunking where appropriate. As educators we should also be aware that overloading the working memory to long term memory structure is very counter productive.

If learning is measured by the ability to retrieve stored memories then we should also look at how best to ensure that retrieval is possible. This usually requires a conscious processing of the information to have occurred at least once. Having been through the processing in working memory information will be in long term memory. That does not necessarily mean that the pathways to it will be strong. The more often we retrieve a given memory the strong the pathway to it becomes. So, for learning this means there is a need to revisit information in order to “solidify” the ability to retrieve it easily. Simple really – revision works.

It is also fairly clear that linking new knowledge to existing knowledge is helpful through the development of knowledge schema and can help build memory pathways. This is not necessarily the same as saying that the learning of new knowledge depends on the existence of previous knowledge. Just that its retrieval is facilitated by being linked to existing knowledge in a schema. This is as much a basis for thematic learning as it is for core knowledge. So as educators it makes sense to rehearse connected knowledge with students prior to initiating new learning. As students it makes sense to create narratives that link knowledge together.

As a complete aside it is interesting to note that a lot of the work that is carried out building computer-based neural networks used the computer language LISP, which is a high level list processing language.

Based on our understanding of the way the brain works there are a few simple skills we should adopt, as learners and as teachers to assist learning. Beyond that we need to take care. We are currently at a point with our knowledge of how the brain works where what seeps from the research labs into the real world is accessible enough to make people think they understand it and can use it to make predictions, that we can say with certainty “this works” or “this doesn’t work”. In reality, save for a few simple matters alluded to above, we are a long way from there. Much of the research is carried out on such small groups the results are often only useful in that they suggest further research would be useful. Most of the rest of the “brain” related advice tends to be more of the psychological research which, interesting though it is, may well be useful when applied to populations, but is not always relevant for any given individual.

My advice would be to take with a large pinch of salt any advice about how we learn given on the basis of what we “know” about the brain. And that probably includes this!

Cognitive Function Diagram (c) Cisco Systems 2008