I’d rather not spend time on this if anyone knows the answer.
L5 corticothalamic cells definitely project to the brainstem and such, but that doesn’t mean they’re motor outputs. Many corticofugal projections seem to target sensory cells, modifying sensory input rather than generating behavior.
S1 seems to activate a muscle (source), so motor output still seems universal. I’m just wondering whether motor command cells are corticothalamic.
I would like to know the answer to this as well. I assume this is so, but my research was inconclusive. It also seemed necessary for cerebellum to at least modulate the corticothalamic output it since fine tuning is required when it comes to any sort of movement.
Can you elaborate on what you’re saying about the cerebellum? I think it could influence cortically generated behaviors because it projects to cortex via thalamus. Although, I don’t know whether it could influence motor outputs from things besides the motor cortex.
Let’s say the layer 5 sends the motor command for a specific movement, what happens if that command gets executed insufficiently or the desired outcome is not quite reached? Then you should either modulate the corresponding layer 5 to steer into the desired outcome or you can keep the movement command seperated from its execution (stuff that goes wrong specifically). There could be countless things that could go wrong during the execution of a single movement; external factors, injuries, physical capability at that given time etc. It would be a combinatorial explosion if you solve this by modulating the layer 5. There is also the process of learning that said movement which (I think) needs some sort of stable motor command to use it as reference to correct on. So from what I have read, the execution of the said movement is constantly corrected and fine tuned (possibly optimized) to allign with the actual wanted movement via stuff like balance, locomotion constraints. Cerebellum specifically does this.
I think the preparation is also something which could cause a movement mistake. For example, in a movement sequence, it should do each step in a way which prepares for the next one, like right momentum, positioning, balance, or whatnot.
Preparation can also cause a movement mistake when doing just one movement. There’s preparatory activity even for a single movement, even in the spinal cord. The preparation might involve actual motor activity, like balance or holding an arm in a particular position, I don’t know.
(Btw, preparatory activity in spinal cord means even if L5ct cells project to spinal cord motor neurons or premotor neurons, they still might not send motor commands.)
So does there need to be a separation between defining the desired movement and executing it? For novel behavior patterns generated by cortex, the cortex is the one saying what the desired movement is.
For example, let’s say the cortex wants to pick up a cup. Would the cortex tell the cerebellum, I’m gonna pick up this cup (so sensory / goal info, not just coding movement), then do that while the cerebellum corrects any accumulating mistakes during movement?
Those are definately good examples.
I was thinking more along the lines of cortex outputting the command (movement is coded, not sure whether it needs the rest) to pick the cup up but your hand can be a bit shoddy on executing the necessary action or sequence of actions when there is stuff going on like momentum, balance etc. as you said. In short, there is a matter of accuracy during the movement. Sometimes it is better and sometimes it is worse even if the command was the same. Execution is never exactly the same. It also gets better in time with practice. So yes, I think there needs to be a seperation to solve this efficiently which is the point of cerebellum.
On the other hand, this does not really answer the initial question. It only allows me to speculate that it would be less efficient to directly modulate the corresponding layer 5 activation (changing the command) and more efficient to modulate the corticothalamic projection (tuning the execution without changing command).
According to the article below the humans’ spinal cord may have potential to fine tune certain motor movements. Taking it to the extreme there’s the octopus with each of its tentacles having its own tiny brains possibly for movement control. May not be what you’re looking for but quite interesting nevertheless
That makes sense.
Cerebellum would need to modulate subcortical motor things, I think. The corticothalamic projection just impacts inputs to cortex.
I think the distinction between motor commands and execution is pretty useful for my question. I wasn’t sure whether the brain distinguishes between motor copy signals and self-movement signals. Now I think that has to be true (for somatosensation at least. Maybe not vision.)
The corticothalamic projection is often interpreted as a motor copy signal. It’s for reference frame transformations, for example so you don’t see the world move when you move your eyes. What the cortex actually needs is a self-movement signal. That makes me expect that L5 corticothalamic cells don’t generate movement.
The barrel cortex seems to do a reference frame transformation using a self-movement signal from the thalamus (source1, source2, source3, source4, studies on self-movement coding in the whisker system).