How Brains Race to Cancel Errant Movements

By Neuroskeptic | August 3, 2013 3:02 pm

You’re just about to put your hand on the hob of your electric cooker, when you remember that it was on full blast until five minutes ago, and will still be scalding hot. You try to stop yourself – but will you succeed?

This kind of ‘stop!’ scenario is the subject of some most interesting work by Ann Arbor neuroscientists Robert Schmidt and colleagues: Canceling actions involves a race between basal ganglia pathways

Stop intentions may or may not succeed. Psychologists study these using ‘stop-signal’ tasks in which, for example, you must press a button whenever you hear a ‘go’ signal, unless it is followed by a ‘stop’ signal.

Stop-signal task performance is well described by theoretical models in which the Go and Stop cues respectively initiate stochastic go and stop processes that race for completion. The outcome of this race condition determines whether stopping is successful.

Until now however, this ‘race’ has been purely conceptual. Schmidt et al present data showing that a physical race between two neural pathways is at work.

Lab rats were trained on a stop-signal task. They had to move their nose (or not move it, on stop trials) in response to sounds, in order to get food. Meanwhile, electrodes were recording neural activity in a number of brain areas: here’s Schmidt et al’s cartoon diagram of the relevant circuitry.

It turned out that, in response to ‘stop’ signals, activity in the subthalamic nucleus (STN) spiked up. However, this happened regardess of whether the rat succesfully avoided moving.

In the substantia nigra (SNr), however, activation in response to the ‘stop’ signal only happened on trials where the animal ended up not moving.

In other words, on failed stop trials, the brain’s ‘stop’ command went missing, somewhere in the post between the STN and the SNr. It was sent, but not read.


Because another area, the striatum, got there first. The striatum, responsible for relaying ‘go’ signals, projects to the SNr but it uses GABA, an inhibitory neurotransmitter which is able to, in essence, block out the ‘stop’ signals, but only if it gets there first.

Further analysis revealed that when the striatal response to the initial ‘go’ beep was fast, this corresponded to a fast movement but also to a failed stop signal (if there was one).

Slower striatal responses, however, corresponded to slow responses and succesful stops. Being slow, they were not fast enough to out-race the stop signal (if any).

The strength of the striatal ‘go’ response was quite irrelevant – it was all about the timing. And the differences were of the order of 1/10th of a second:

The speed of the stop signals was very reliable. What varied from trial to trial was time it took for the ‘go’ signal.

But why? It’s curious because the ‘go’ beep was always the same. So was the response required of the rat. Why did the rat’s striatum sometimes take longer to do the exact same task?

The speed of signal transmission in nerve cells is fixed by the laws of physics – so that’s not a likely explanation. The striatum’s inconsistent timing must arise at a higher level of organization.

ResearchBlogging.orgSchmidt R, Leventhal DK, Mallet N, Chen F, & Berke JD (2013). Canceling actions involves a race between basal ganglia pathways. Nature Neuroscience, 16 (8), 1118-24 PMID: 23852117

CATEGORIZED UNDER: animals, papers, select, Top Posts
  • AnneWaller57

    The horse race model and the math to test it, and the validation, were all done decades ago by psychologists. Who really cares that it is in the basal ganglia? It had to be somewhere!

    Logan, G. D., Cowan, W. B., & Davis, K. A. (1984). On the ability to
    inhibit simple and choice reaction time responses: a model and a
    method. Journal of Experimental Psychology: Human Perception and Performance, 10(2), 276.

    • Jillian Carson

      I do I have Parkinson’s disease and sometimes we just can’t stop

    • Wouter

      “Who really cares that it is in the basal ganglia? It had to be somewhere!”

      Well, you’re satisfied easily. From your point of view, we can just stop doing science altogether; all the stuff that we’re investigating has got to be somewhere. So who really cares, right?

      If we ever really want to be able to successfully intervene with the brain (e.g. fix certain disorders/diseases), it’s essential to know what does what and how.

    • Neuroskeptic

      Mmm. It had to be somewhere – if the model was correct. But it might have been that the model was an accurate description with predictive power, but which didn’t actually correspond to physical reality (like the solar-system model of atoms).

  • martoiu

    Probably the free will stopped the mice from stopping.

  • Katherine Krein

    For some reason this reminds me of the children’s game: Simon Says:

    Listening to cues and then mirroring the physical motions of the leader.

    The mirroring of the physical motions sometimes gets ahead of processing the verbal cues. You mirror the motion before you realize that the verbal cue was not correct.

  • Bradley Voytek

    I don’t want to “scoop” him, but a collaborator of mine, Mike X. Cohen (, gave a really cool talk about what he’s calling “partial correct” responses in behavioral tasks. The gist is that, for most behavioral experiments, responses are recorded via button presses (which the computer records as binary yes/no responses).

    What he found, however, was that if you actually look at the EMG of the response hand, you’ll see that on many trials where the system reports that the participant correctly made no response, for example, the EMG actually shows that the person initiated a movement to respond when they shouldn’t have, but managed to stop the response before going all the way to pressing the button.

    Anyway, these partial correct responses have wildly different neural patterning (to describe it in a hand-wavy way), which normally get folded into the overall “correct v incorrect” analyses. It’s a really neat, simple finding, and seems like this paper here is highly relevant.

    Because the bigger question is, how the hell is the CNS able to STOP a motor command after it’s already been initiated?! Meaning, the motor system is relatively very fast compared to cognition, and a person has already initiated a movement, then realizes their mistake and halts that movement. That’s pretty incredible.

    • Neuroskeptic

      Well, if these mice are anything to go by, the answer is that the ‘stop’ pathway is inherently fast. Even if the stop signal occurs after the go signal, the stop pathway can win the race against the go pathway except when the go pathway is ‘on top form’ i.e. at the fast end of its reaction time distribution.

  • Cindy

    predictability and habituation

  • Pingback: | Kalex's Tome()

  • Jari Peuhkurinen

    Is this same process as indecisive action? What I mean is that if a person needs to make a decision to go forward against some type of treat, that we would normally want to run away from. After the decision to go forward our brain tries to cancel the decision. Persons actions may seem indecisive.

  • Nicolas Blackburn

    « Because another area, the striatum, got there first. The striatum,
    responsible for relaying ‘go’ signals, projects to the SNr but it uses
    GABA, an inhibitory neurotransmitter which is able to, in essence, block
    out the ‘stop’ signals, but only if it gets there first. »

    Interesting. The necessity of inhibitory neurotransmitter seems natural otherwise the brain would receive two influx of contradictory signals and could therefore be probably impaired to function normally. This is analogous to file locking on computer file systems to avoid the files being accidentally overwritten and thus become corrupt. This seems to indicate that locking behavior occurs naturally in biological concurrency control systems.

    I’m curious to know if the purpose of this kind of locking in biological systems is the same as in computer systems, that is to avoid some kind of information corruption critical to the normal functioning of the general system.



No brain. No gain.

About Neuroskeptic

Neuroskeptic is a British neuroscientist who takes a skeptical look at his own field, and beyond. His blog offers a look at the latest developments in neuroscience, psychiatry and psychology through a critical lens.


See More

@Neuro_Skeptic on Twitter


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar