Dispelling myths about Dopamine. The Chemistry of Error and Motivation
Dopamine is in the cultural zeitgeist. Want to lose weight? Eat the Dopamine Diet. Want to dress happy? Try Dopamine Dressing. Want to be a Silicon Valley hi-tech entrepreneur genius? Boost your dopamine. Want to feel constant pleasure? Stuff as much dopamine in your noggin as you can.
Craving that dopamine hit? Then check your phone. Read a story. Watch an advert. Or visualise something good happening.
Dopamine will enhance your mental prowess, give you the figure you always dreamed of, and make you happy. Not bad for a lowly chemical sloshing around in your brain. And all of it utter rubbish.
Here’s what it does. Dopamine either changes fast or slow. Fast changes are an error signal. Slow changes are a motivation signal. That’s it.
Dopamine: fast and slow. Error and motivation. These correspond to two different ways dopamine is released. One way is a precise, short, large spike in the amount of dopamine, in a small region of brain. The other way is all the time, creating a constant, low concentration soup of dopamine sitting around in many regions of your brain.
(A couple of things to get out of the way. Dopamine is in many places in the brain. There’s some in your eyeball, for example. But when the media says “dopamine” it means the dopamine releasing neurons in small groups in the middle of the brain. And they also mean where those neurons release their dopamine: a big brain region tucked up nice and warm just under your cortex: the striatum).
How is that precise, short, large spike an error signal? Say you wandered into my house unannounced and, instead of throwing you out on your ear, I offered you a chocolate biscuit (McVities, obviously). Your dopamine neurons would burst into life, spiking dopamine. They signalled the error between what you predicted (being forcibly ejected with a hoof to the bum) and what you received (a nice biscuit). This prediction error was in your favour – it was a positive error.
Say I asked you to turn up at my house at 3 o’clock so I could you give you a chocolate biscuit. You turn up at 3 o’clock, and I give you the promised biscuit. What do your dopamine neurons do? Sod all. You predicted you’d get a biscuit at 3 o’clock, you got a biscuit at 3 o’clock; all is right with the world. Nothing surprising is going on. There was no error.
What if, when you turned up at 3 o’clock, I didn’t give you that chocolate biscuit? What if I just blithely ignored your presence instead? Then your dopamine neurons would briefly pause their activity, stopping the release of dopamine. They signalled the error between what you predicted (a chocolate biscuit) and what you received (nothing). This prediction error was not in your favour – it was a negative error.
This is what fast dopamine does: it signals the error between what you predicted and what you got. That error can be positive, negative, or zero.
It is not reward. Dopamine neurons do not fire when you get something good. They fire when you get something unexpected. And they sulk when you don’t get something you expected. Rewards make you happy. Dopamine does not.
And that slow dopamine? The slowly changing, but always low, concentration soup of dopamine? It signals how motivated you are to work. The higher the concentration of this dopamine soup, the harder rats work for food. Take that dopamine soup right down, and rats won’t budge off their backsides to work for food. Dopamine soup is for wanting; is for signalling to your neurons: commit to this course of action. Not happiness.
(Notice how short that paragraph is compared to the bit about the fast, large spike of dopamine? That’s because we have a detailed theory about the fast, large spike of dopamine, a computational theory that both makes sense of what we see in experiments, and makes real, testable predictions for future experiments. It is one of the very few computational theories we have about any part of the brain. Indeed, it’s so unique that just last week the one million euro Brain Prize was awarded to three of the people who did much to trtablish that theory. The slowly changing soup is in the same category as practically all other studied parts of the brain: we have some ideas, but a broadly agreed theory is a long, long way off.)
There’s no getting away from the fact that the way dopamine works in the brain is complex. We know that this fast and slow split is not as simple as we’d like. For example, it seems there are some dopamine neurons that use their fast firing to affect movement. On the one hand, this is potentially fantastic, as it helps make sense of why the loss of dopamine neurons in Parkinson’s disease causes movement problems. On the other, it does not fit with an error signal. It seems that the error signal is carried by a separate set of dopamine neurons.
And there’s no getting away from the fact that there’s nothing special about dopamine. It’s on a list of tens, likely hundreds, of different chemicals that the brain uses to convey information from one part to another. There are only a handful of chemicals it uses to communicate directly between neurons, where the activity in one neuron causes an immediately measurable effect in another. But there are a huge number, including dopamine, that change that communication. That make it easier or harder; slower or faster.
Serotonin, norephinephrine, enkephalin, substance P (a real thing), neuropeptide Y (another real thing), mu-opioids (a more plausible candidate for happiness): the list is endless. There’s a simple circuit of neurons in the lobster’s stomach that controls part of its digestive system. Its neurons use one of two chemicals to directly communicate. But it uses twenty-one chemicals to alter this communication.
There are so many other brain chemicals to chose from, we can guarantee the current fixation on dopamine is just a fad. Mark my words, soon it will be neuropeptide Y: NPY Diet! NPY Dressing! NPY Brain Training! Many more crimes are yet to be committed against poor defenseless brain chemicals.
What is the root of these crimes against dopamine? A prime suspect is that, in trying to explain what it does, writers have chosen to leave out the wrong words. The main research papers on dopamine call its fast, large spike (and its fast, sulking pause) the “reward prediction error system”. A bit of a mouthful; and hard to remember. So textbooks and popular science articles shorten this to the easier “reward system”. Even the esteemed journal Nature, when reporting on the Brain Prize, called dopamine the “reward system”. But it is not. “Reward” is precisely the word they should have left out. We can replace the word “reward” with the word “outcome”, and it makes not one jot of difference to the theory. Yet by calling it the “reward system”, the mental leaps to happiness, pleasure, emotion and all the other crimes are easy ones to make. No, it is the other words – “prediction” and “error” – that are the key.
Dopamine is a prediction error system.
So when a friend or neighbour, relative or co-worker, lifestyle magazine or proud scientific journal commits a crime against dopamine, gently whisper in their ear “prediction error”. And, yes, they will move a little bit away from you. And, yes, they may well feign a sudden need to check whether they left something in the car. But you will be a hero to science.