"So dance-dance like it's the last-last night of your life-life; gon' get you right."

In previous entries, I've described the brain systems that drive people to dance and allow them to physically do it. There is another essential process in this equation, though, and that is learning. Dancing is definitely a learned motor skill, and beyond learning the basics of how to perform various moves, there are many more ephemeral aspects to learn. These range from dance etiquette and customs, to identifying dances and matching them to music, to style considerations. The topic of generalized learning is far too broad to cover in a single blog entry, so I will focus on motor learning.

When we look at which parts of the brain help us learn dance moves, we find our old friends, the cerebellum and basal ganglia! "Motor learning" refers to the process of becoming proficient in a given physical movement, and in the case of dancing, there is technically a correct way to perform any dance move. This type of learning, in which the current attempt is repeatedly compared with a model showing the correct method, is called "supervised learning." Though scientists haven't exactly reached consensus on this point, one prominent theory claims that the cerebellum is specialized for performing supervised learning: the cerebral cortex provides the model, and the circuitry of the cerebellum compares current motor behavior with the model and tries to change behavior when it differs from what the model says is "correct."

Of course, this view of the cerebellum assumes that the person already has a mental representation of what they should be doing. The questions of how this representation is formed and what it looks like are difficult to answer, but the basal ganglia seem to be involved. Remember that dopamine levels in the basal ganglia can either cause a planned motor behavior to start, or can stop it from happening. The cerebellum can't refine a set of actions if the basal ganglia don't let the body even start acting--in this way, dopamine levels serve as a measure of which actions are worth learning to begin with. If a dancer thinks a certain dance move is really cool, then that move might be associated with high levels of dopamine in his or her basal ganglia. When the dancer's cortex proposes trying that move, the basal ganglia will be flooded with dopamine and will probably reinforce that behavior, allowing it to begin. Once the dancer starts moving, the cerebellum will try to correct muscle mistakes and make the actual move look like the dancer's mental representation of what it should be. This is the essence of practicing physical activities in order to learn them.

I haven't yet addressed how this mental representation might be formed or evaluated. This is probably the job of the cerebral cortex, and especially the prefrontal cortex. These brain regions are responsible for so-called "higher mental functions" such as attention, working memory, planning, and decision-making, and are not as straightforward as the basal ganglia or cerebellum. I will just mention that, in the theory that assigns supervised learning to the cerebellum, the cerebral cortex is said to perform "unsupervised learning." This essentially means that there is no explicit "right answer." The system observes the world and tries to find patterns or interesting regularities among the chaotic inputs. There's no model to check against except the outside world itself.

While unsupervised learning can be slower and less reliable than supervised learning, it is also much more versatile. To give just one example, infants learn their first language through unsupervised learning. A dancer trying to learn a new move by watching someone else perform it is engaging in unsupervised learning. Both types of learning are crucial in all aspects of life, from the most fundamental survival tasks to the highest expressions of human creativity--and they each depend on specialized collections of neurons within the brain. This is why I study neuroscience.

(photo © dmason for openphoto.net CC:Attribution-ShareAlike)

"Just dance--gonna be okay."

Modern neuroscience suggests that the basal ganglia are responsible for initiating motor actions, based largely on the amount of dopamine associated with each hypothetical movement. Starting a motion is obviously only the first step, though. Some brain region must also control the motor action as it occurs, and given that healthy humans very readily adjust their movements at all stages to keep them accurate, this motor control seems to be an important task. Examples are everywhere: just imagine reaching for an apple on a tree on a slightly windy day. The goal of your action may move through space quite a bit as the branch sways in the breeze, but you don't have to start over each time; when you see the apple move, your hand continues toward the apple's new position without any conscious effort from you.

As it turns out, motor control is linked to a distinct part of the brain called the cerebellum. Though it looks like a small afterthought stuck right behind the brainstem, the cerebellum actually contains far more neurons than the rest of the brain combined. These neurons are arranged into a few compact circuits that actually fit the computer metaphor better than does most of the brain. In some ways, the main difference between different parts of the cerebellum is simply where their inputs come from and where their outputs go: the computations performed between input and output are very similar.

Many of these inputs come from regions of the brain in the cerebral cortex that generate motor actions. The main function of the cerebellum in this case can be seen as implementing plans for movement: the motor cortex proposes a plan of action, the basal ganglia decide whether it's worthwhile, and if so, the cerebellum makes sure that the body follows the plan accurately. As with the basal ganglia, many of these insights into function come from cases where people have suffered damage to their cerebellum. Often, such patients can initiate actions, but they are very uncoordinated and clumsy while performing them. This is in contrast with basal ganglia damage, which can lead to patients unable to start voluntary movements, but capable of carrying out actions they've already begun.

Dancing is a combination of broad, whole-body movements with finely detailed movements, and as such, any dancer owes a great deal to their cerebellum! As humans, the true complexity of our everyday actions is mostly invisible to us: we simply act, and the extensive neural processing required stays hidden from our consciousness. Discovering the full inner workings of our brains is a difficult task, but I believe it will lead to insights into our behavior that make it worthwhile. As one example, damage to the cerebellum can cause symptoms that look like simple drunkenness. Investigating the cerebellum may help us understand how alcohol affects the brain, which could lead to new approaches to prevent drunk-driving accidents. Specific applications like this would be easier to develop with a more complete general understanding of the brain.

(photo © Jasenka Petanjek for openphoto.net CC:Attribution-ShareAlike)

"Should I stay or should I go?"

I've mentioned that dopamine drives motivation, and that more dopamine tends to reinforce actions. The brain isn't just a vat full of chemicals sloshing around, though: most of the space is taken up by brain cells. These neurons come in different shapes, but the basic mechanism is almost always the same: one end takes inputs, the other end sends output, and the parts in between decide whether or not to send an output based solely on the sum of the inputs.

This may seem too simple to account for thought and behavior, and indeed, most of the interesting behavior produced by the brain comes from the ways that neurons connect to one another. Neurons connect in complex patterns that are fairly consistent over time and across different people: if you scan my brain, your brain, and the brain of some random person on the street, you'll find the same basic connections in the same places. These connections can be thought of as computer circuits in some ways, and different parts of the brain have circuits that perform different tasks. Since people tend to develop the same circuits, we can examine them and figure out what they do.

Basal Ganglia circuit. Copyright Andrew Gillies.

One of the neural circuits relevant to dance is called the basal ganglia. It's the home of dopamine, where most of it is stored and released. It's also crucial for control of physical movement: for example, Parkinson's disease occurs when just one small part of the basal ganglia dies. This isn't a coincidence. As it turns out, dopamine levels in the basal ganglia determine whether we start a planned physical action or not: high dopamine levels lead us to act and carry out the current plan, while low dopamine levels stop us from acting, and may lead us to come up with a new motor plan that releases more dopamine. Parkinson's patients lose their dopamine-releasing neurons, so their constant low levels of dopamine leave them unable to act on any of their thoughts and plans.

It seems, then, that dance owes its popularity to this circuit on two different levels. On one level, dancing is a motor activity, and relies on this deep system of neurons for carrying out the intricate and detailed movements required. On another, more fundamental level, the decision to dance may be made at least partly by the balance of dopamine and the ways it affects the flow of signals in the basal ganglia. If oxytocin causes pleasant feelings and leads to more dopamine release during dance, then the basal ganglia make sure that the surge of dopamine leads to more dancing!

"No regrets, just love. We can dance until we die; you and I will be young forever."

Last time, I wrote about how dopamine has kept me coming back to the dance floor, by associating dance in my brain with some of its rewards. This raises the question: what are the rewards of dancing? This is a complex topic with different answers for different people, but right now I'm interested in one factor that seems fairly common: oxytocin.



Oxytocin is one of several hormones released in small amounts during sustained close contact, as in dance. Its effects tend to build implicit trust and spur further close contact--it has been called the "love hormone." This can be illustrated by looking at the results of two experiments. The first involves prairie voles. In the wild, these small mammals mate for life, but only after spending a lot of time together. When given oxytocin before their first meeting, a male-female pair of voles will skip the "getting to know each other" stage and proceed right into wild and enthusiastic pair bonding, mating with each other continuously for a whole day and then remaining monogamous as normal. This seems to indicate that oxytocin is an important factor in building trust and familiarity through sustained contact.

The second experiment involves humans, in a less torrid setting: participants play a game in which they decide how much money to invest with another person, knowing that the other person ultimately decides how much money to give back. Those people who were given a nasal spray of oxytocin beforehand consistently invest more than the control group who were not given the hormone. When the human interaction was replaced with a computer interface, oxytocin made no difference in the amount invested. This suggests that oxytocin increases trust in humans as well. Read more about this experiment here.

Due to the low acceptability of social touch in modern American culture, dance's potential as a catalyst for romance or strong friendship is especially prominent. One of the first stumbling blocks for many new dancers is learning to be comfortable touching their partner. Often, all it takes is just doing it--take someone's hand, dance to a song, and let the oxytocin flow. As for what happens after that, well, I don't have any scientific research to share, but an informal look at the romances that have blossomed within the ballroom dance club is very suggestive. The details differ, but we may have a lot in common with those prairie voles when it comes to social bonding!

Why do brains dance?

Image from my-ballroom-dance-shoes.com

There are many reasons that people might choose to learn ballroom dancing. It is a highly social, physical activity, with the added potential for creative expression. These are all fine reasons, but primarily, I joined the thriving ballroom dance club at UCSD to meet girls.

As a cognitive neuroscience major here at UCSD, I don't find this surprising. Like most processes in the brain, the strong biological drives focusing on romantic partnership (and some of its associates) involve the release of various chemicals called neurotransmitters, which go on to generate meaningful electrical activity in the brain. The intensity of this response can be increased by the close contact and physical exertion experienced during social partner dances.

To name one famous example, dopamine is an important modulatory neurotransmitter. This means that it modifies the behavior of certain brain circuits in more subtle and indirect ways than other neurotransmitters. In addition to other functions, dopamine serves as a primary agent of the brain's broader system of motivation. Experiments have found that dopamine is released into certain parts of the brain when an unexpected reward is received, and the effects of this release tend to promote repetition of whichever recent action seems most responsible for the reward. Further research shows that as an action is rewarded in the same way repeatedly over time, the corresponding dopamine release tends to occur progressively sooner after the action is performed, rather than the after presentation of the reward itself.

If you'd like to learn more about how dopamine functions, how people research it, and what effects it has on people's lives, Scientific American has an interesting article on the subject: The Dope on Dopamine

As for me, I plan to continue dancing--and blogging about it. Thanks to dopamine, I've associated some very nice social rewards with ballroom dance, and thanks to cognitive science, I know something about why!