Learning isn't all about memory

Learning isn’t all about memory. Not in the traditional sense. Much of our learning involves the adopting of new behavioural routines. Ways of responding to the world based on the context around us. Importantly, these kinds of learning are about preparing our bodies to respond to predictable events.

For example, when one walks a set of stairs, and expects there to be one more stair than there is, when the final stair never materialises, we briefly panic and our body flails about to manage the disorientation. In this case, both the way we walk up stairs and our response to the missing final stair are learned behaviours. Behaviours you’ve added to your action routines that you never had to memorise.

These kinds of routines are the most basic forms of learning, and in fact, form the foundation of our emotions. Emotions aren’t random. Rather, they’re the product of when these learned routines are ‘interrupted’. This explains the moment of panic caused by the missing stair.

¶What isn’t learning

Before we talk more about this kind of learning, it’s important to point out that we don’t need to learn everything. Some behaviours aren’t learnt at all. We can split these behaviours into two rough categories.

¶Reflexes

Reflexes are automatic and involuntary reactions to specific environmental conditions. For example, if you touch something hot, you jerk your hand away. This particular behaviour is never learned and it’s managed by nervous system pathways that are quite complex, but ever-present.

¶Instincts

Instincts are even more complicated behaviours that, like reflexes, are carried out in response to very clearly defined environmental contexts. For example, salmon swimming up river to reproduce, or dogs marking trees, or sleeping.

These behaviours are not learned, but are genetically prescribed. Instincts can include reflexes, but are more commonly used to describe more complicated routines of action sequences.

¶Sensitisation and habituation (noticing and ignoring)

The most simplistic kinds of learning are called sensitisation and habituation. Essentially to ‘notice’ and to ‘ignore’ respectively. We ignore events with no consequence but notice events with important consequences.

When some unexpected event occurs, we will first orient to it. This orienting behaviour is a reflex. Consider, for example, a car alarm. The alarm sounds and you turn to check it out. Then one of two things could happen. If there’s nothing around the shrilling car, we will start to ignore it until eventually it fades into the background. That’s habituation. However, if the event was less innocuous, say for instance someone was poking you with a sharp object, or a baby was crying, rather than ignoring it we’d attend more and more strongly to it until the event is over. This is sensitisation.

¶Classical conditioning

We must also learn what consequences predictable events might signal. Events that happen often have predictable outcomes that we can learn. This means we can prepare our bodies to respond automatically. For example: looking at a lemon, or thinking of one, makes our mouths water. We have learned to prepare for a sour taste and our mouth is preparing accordingly.

Typically, this kind of learning consists of reflexes/involuntary responses being paired with events we come across frequently in the worlds. We learn that the new event (the appearance of a lemon) predicts an event that originally caused the reflex (a sour taste in the mouth), so we prepare to respond (our mouth waters).

¶The bizarre origins of classical conditioning

Ivan Pavlov is psychological legend. He was very excitingly working on the digestive enzymes in saliva and stumbled across what would become an entire branch of psychology. He was collecting saliva samples by spraying powdered meat at restrained dogs. Pavlov noticed that, after a while, the dogs would start salivating as he was restraining them, long before he turned on the powdered meat hose (or whatever one uses to spray dogs with powdered meat).

This was the discovery of what’s now known as classical conditioning. Pavlov recognised that the dogs were anticipating the meat powder because of the association between the meat powder and the restraints. He did some further experimenting with bells, and found that he could (by pairing the meat powder and the bell) get the dogs to drool again regardless of whether the meat powder showed up or not, just like with the restraints.

Eric Kandel separately (and much later than Pavlov) was doing some work on aplysias—a kind of sea slug. He found that he could identify this pairing of seemingly unrelated things in the nerve cells of his aplysias. See, these slugs, like us, have a withdrawal reflex, and by pairing a poke in the gills with shrimp juice he could get the slug to ‘learn’ to withdraw when coming across the shrimp juice.

This kind of learning is, then, subconscious. Think of all the times you’ve thought about something, and another thought has popped into your head unbidden. Or smelled something, and remembered another time you smelled that same thing. We don’t need to actively make connections between related things in our minds, when these relationships are predictable. This kind of learning saves us work.

¶The errata of memoryless memory

There are caveats however:

• We learn some things more easily than others, for example:
• The more noticeable the stimulus, the easier we learn, and the more likely it is to dominate other, weaker stimuli.
• The closer together the pairing, the quicker we learn.
• If the stimulus comes after the thing you are trying to pair it with, you won’t learn the relationship.
• Some stimuli seem to actually be something we’re primed for from birth. We learn much faster when we pair things with certain stimuli like snakes or spiders – a phenomenon thought to be responsible for some of our phobias.
• If the pairing doesn’t happen for a while, you’ll lose the automatic response (called extinguishing)
• Sometimes, even when a response has extinguished, it’ll happen randomly sometimes (called spontaneous recovery)
• Once you’ve conditioned two things together (like a noise and food, so the noise now makes you salivate), you can use the noise to condition you to salivate to another, previously unconnected stimulus. This is called sensory preconditioning.

¶Outro

Classical conditioning describes an enormous number of the intricacies of this predictive quality within us. But the critical factor is that we don’t always use memory to learn. So long as two things have a clear and predictable relationship, we will learn to respond to them in the same way.