You can’t learn something unless you almost know it already. Decades ago, that principle was known as Martin’s law. There’s a corollary: It’s harder to learn something strange than something familiar; or, new learning builds on old learning.
Associative learning is connecting the dots—or the synapses, in many accounts at varying levels of sophistication, following Donald Hebb’s insight into how learning builds structure among neurons and structure determines what is acceptable for learning.
Associative learning is the business of the classroom and synaptic plasticity is what we all experience there. What is rarely mentioned in popular psychology—though the research literature is exploding with the topic—is the contribution of glia to learning.
Discussions have been about three types of glia in the brain and enteric nervous system: astrocytes, oligodendrocytes and microglia. You can catch the flavor of the discussions here. There are other kinds of glia and a variety of neuron types, but we will concentrate on these three types.
BIO: We are dealing with three types of cells: astrocytes, oligodendrocytes, and microglia. Before tackling their connections to learning, we ought to be sure we can tell them apart. Is it even clear how they differ from neurons? More than one psychologist has undertaken research into the effects of experience on neuron proliferation with hazy ideas about how to distinguish neurons from glia using a common Nissl stain like thionin, cresyl violet, or toluidine blue. But, using a variety of techniques, the distinctions are unavoidable. To an anatomist, at least.
Though still a bit controversial, the principle that learning is based on synaptic changes has been widely discussed in behavioral neuroscience. The changes reflect synaptogenesis as well as synaptic pruning. This is what is meant by synaptic plasticity, although the importance of myelin plasticity also has to be considered.
Astrocytes control synapse formation that is the basis for learning. They secrete proteins at the site of synapse formation that foster synaptic growth. Since a single astrocyte may contact several thousand nearby neurons, astrocytes have a potential ability to link many neurons into networks that learn.
Microglia are the smallest glial cells and the only glia that belong to our immune systems. In addition to ingesting injured cell fragments in the brain, microglia are responsible for the synaptic pruning that contributes to forgetting, which improves the efficiency of learning.
Oligodendrocytes may be the most familiar and most neglected of the brain’s glia, at least when it comes to learning. Like their counterparts in the peripheral nervous system called Schwann cells, oligodendrocytes lay down the myelin sheath that speeds the propagation of action potentials along axons.
The brain is an intricate timing system, and the speed of neural transmission determines the timing of neural events. When a group of active neurons form an engram, the heightened neural activity increases the amount of myelin formed by oligodendrocytes along the neuronal axons and enhances the transmission of future action potentials. As you would expect, injury to the added myelin can cause severe impairment of behavior.
As textbooks catch up with research of the past decade it will become clear to a wider audience that myelin is not packed like fiberglass insulation into the brain’s attic. Through myelin plasticity, it responds dynamically to our experience.
PSYCHO: Fortunately, myths about memory are few. Nevertheless, the neural details of learning are important because behavior is changing all the time, and much of the change is not really learning: sensory adaptation, adjusting sleep to rotating shifts, code switching in bilingual speakers and attentional switching in all of us, gaze following, and so on—all of the changes that adjust our behavior to the environment without making an enduring difference—are nowadays covered by the catchall term “plasticity”, and a common name misleadingly suggests commonality of mechanism.
A decade ago, Vaughan Bell tried to preserve a limited sense of the word, pointing out that generalization had robbed the term of precision. There is in fact a broad range of useful applications, but simply saying the brain is plastic ends up explaining very little. Rather, it implies a general but false capacity of neurons to change for almost any reason.
Updating the story of learning by adding glia to the explanation should rescue us from magical neurons. Glia evolved with a supporting role in the brain, subject to natural selection and other biological constraints, that can be compared to the evolution of learning and memory itself.
Furthermore, the engram appears to be a valid construct with a definable cellular and molecular structure, despite Lashley’s failure to find it. Researchers are building a coherent picture of a manifestly imperfect mechanism.
It will be a while before anyone can detail the neuroscience of a further burden of learning, which is making new learning fit in with everything else one knows, sometimes called system consolidation, which is another aspect of Martin’s law. You can’t learn something unless you almost know it already.
SOCIAL: Participation in groups rests on several kinds of learning provided by observation, language, and . The neural underpinnings of such learning have been partly revealed in the so-called social brain, but the glial contribution is barely understood.
Nevertheless, glial responses to social stress, which is a motive for social learning, has been investigated to some extent. Intraspecific aggression made the myelinated segments of axons shorter and thinner (research here) in the medial prefrontal cortex of mice.
Microglia were reported to have been activated in a proinflammatory sequence of responses to repeated social defeat in mice.
Astrocytes, also, are known to respond to stress, if not to social stressors. Exposure to fox urine, a stressor for mice, can cause retraction of processes of Bergmann glia in the cerebellum (research here) to disrupt communication between astrocytes and neurons.