sciencebriefs
13:00in productionCh. 1 · A lever a rat won't stop pressing/ 13:00 · ceiling 15 min
Neuroscience

Reward system

Dopamine was once billed as the brain's pleasure chemical, on the strength of rats pressing a lever thousands of times an hour for brain stimulation — but closer study now separates wanting something from liking it, and puts dopamine mostly on the wanting side.

In 1954 James Olds and Peter Milner found that rats would press a lever compulsively to receive electrical stimulation of certain brain regions, establishing that the brain has a dedicated reward system rather than reward being a side effect of survival behaviour. Dopamine, released along a pathway running from the midbrain to the nucleus accumbens, was identified as central to that system and for years was treated as the brain's pleasure chemical. Later work, particularly Robinson and Berridge's incentive salience model, split reward into wanting and liking and argued dopamine tracks the first far more reliably than the second, a distinction with direct bearing on addiction, Parkinson's disease and schizophrenia.

Chapters & takeaways6
  1. 0:08
    A lever a rat won't stop pressing

    Olds and Milner's 1954 self-stimulation experiments showed the brain has reward circuitry that can be triggered directly, without food, water or any survival need.

  2. 2:10
    One pathway, one chemical

    Dopamine released from the midbrain into the nucleus accumbens became the leading candidate for what that circuitry runs on.

  3. 4:20
    Wanting split from liking

    Robinson and Berridge's incentive salience model argued that desiring something and enjoying it are separable processes handled by different systems.

  4. 6:30
    The pleasure chemical, demoted

    Blocking dopamine did not remove animals' facial signs of enjoying a sweet taste, even as it removed their motivation to seek it out.

  5. 8:40
    A signal for surprise, not satisfaction

    Dopamine neurons fire more to an unexpected reward than an anticipated one, behaviour better described as tracking prediction error than pleasure.

  6. 10:50
    Where the theory still moves

    A 2019 study manipulating dopamine directly in humans found it did shift how much pleasure they reported from music, muddying the wanting-liking split again.

Worth your time?

Yes. Study the whole thing.

4/ 5
What works
  • the self-stimulation experiments are a genuinely striking starting point, easy to picture and hard to argue with
  • the wanting versus liking distinction gives real explanatory power over why addicts keep using drugs they no longer enjoy
  • the material is honest about where dopamine's role is still disputed rather than presenting a finished theory
What does not
  • the sources cite Roy Wise's research by reference rather than narrating his biography or the dates of his key claims
  • the 2019 finding on dopamine and musical pleasure sits awkwardly against the wanting-liking split and is not fully reconciled with it here
Study it if
  • anyone who has heard dopamine called the pleasure chemical and wants to know why that phrase is now considered too simple
  • readers interested in the neuroscience underlying addiction, Parkinson's disease or antipsychotic medication
  • anyone drawn to a scientific idea that keeps getting refined rather than settled
Skip it if
  • readers wanting a single tidy answer to what dopamine does
  • anyone looking for the biographies of the individual researchers rather than the science itself
The written brief4 min read

A lever a rat won’t stop pressing

The claim under examination is that the brain contains a dedicated reward system, not merely circuitry that happens to respond to food or safety, and that dopamine is its central signalling chemical. The founding evidence came from James Olds and Peter Milner in 1954, who found that rats with electrodes placed in certain brain regions would press a lever hundreds or thousands of times an hour purely to receive electrical stimulation there, with no food, water or other survival benefit attached. That behaviour identified a reward system separate from any particular need it might normally serve. Dopamine, released along a pathway running from the ventral tegmental area to the nucleus accumbens during both natural rewards and this artificial stimulation, became the chemical most closely associated with that system, and for a period was described plainly as the brain’s pleasure chemical.

One pathway, one chemical

Much of the supporting work has used direct manipulation of the dopamine system rather than passive observation. Researchers including Roy Wise studied self-stimulation and drug reward directly, arguing that rewarding drugs and electrical stimulation act on central reward mechanisms more powerfully than natural rewards because they reach the circuit directly rather than through the sensory pathways an animal would normally use to detect food or a mate. Later studies used microelectrode recordings of individual dopamine neurons during behavioural tasks, and brain imaging in humans, including a 1998 study using a radioactively labelled tracer to track dopamine release during goal-directed movement. Robinson and Berridge’s 1993 incentive salience model reframed the whole picture, proposing that wanting and liking are handled by separable systems and testing that split with measures such as animals’ facial reactions to taste.

Wanting split from liking

What has held up is that dopamine tracks motivation and the pursuit of reward with real reliability. Dopamine-depleted animals lose the drive to seek out food even when their taste reactions to that same food, once it is placed directly in their mouths, look unchanged, a dissociation that supports treating wanting and liking as separate processes. Dopamine neurons have also been shown to respond more strongly to unexpected rewards than to fully anticipated ones, and to drop below their normal background activity when an expected reward fails to arrive, a pattern that fits a signal tracking the gap between expectation and outcome rather than pleasure itself. This prediction-error account, developed through models such as temporal difference learning, has become a well-supported description of what dopamine neurons are actually doing moment to moment.

The pleasure chemical, demoted

What has not held up is the simple claim that dopamine is pleasure. Blocking dopamine transmission with drugs reduces motivation and can produce anhedonia, a reduced capacity to feel pleasure, but studies measuring animals’ immediate facial and behavioural reactions to a sweet taste found those reactions largely unchanged even when dopamine signalling was disrupted, suggesting the felt enjoyment of a reward and the drive to seek it out are not the same process after all. That picture was complicated again by a 2019 study in humans that raised and lowered dopamine levels pharmacologically and found reported pleasure from music, including the intensity of musical chills, shifted in both directions with it, a result that sits uneasily with a clean separation between dopamine-driven wanting and non-dopamine liking, and has not been fully reconciled with the incentive salience model.

A signal for surprise, not satisfaction

The stakes of getting this right extend well past reward theory. In addiction, the wanting-liking split offers an account of a pattern that otherwise looks paradoxical: users whose desire to take a drug keeps rising even as the pleasure they get from it falls with tolerance, a mismatch straightforwardly explained if dopamine governs wanting rather than liking. In Parkinson’s disease, the loss of dopamine-producing neurons in the substantia nigra explains the movement difficulties that define the condition, while the standard treatment, levodopa, works by supplying the raw material the remaining neurons need to make dopamine themselves. In schizophrenia, most antipsychotic medications work by blocking dopamine receptors, and understanding which parts of the dopamine system relate to motivation rather than pleasure or perception bears directly on why those drugs help some symptoms and blunt others.

Where the theory still moves

Yes, particularly because the subject rewards scepticism rather than punishing it. The lever-pressing rats are a memorable enough image to hang the whole story on, but the real interest is in how thoroughly later researchers took apart the tidy label of pleasure chemical once they had better tools for measuring what animals actually wanted versus what they actually enjoyed. Anyone who has repeated the phrase dopamine hit without thinking much about it will get a clearer, more defensible picture here, including where the current account still strains, as in the 2019 music study. It is a good use of an hour for readers who want their popular neuroscience checked against the more careful, less quotable version underneath it.

Same field · Neuroscience4 of 45
Up next in Science

Focal mechanism

· 13:00

A single earthquake can be read as a diagram that looks like a beach ball. Which quadrants come out shaded tells you whether a fault pulled apart, pushed together or slid sideways — and that reading once settled a live argument about seafloor spreading.

13:00