Terrific new article on motor learning

[jkubie]

Very clever and important article on motor learning.

Motor cortex is required for Learning but not executing a motor skill.

Stereotyped behavior in the videos is spectacular!

summary:
•We train rats to execute spatiotemporally precise task-specific motor sequences
•We show that motor cortex is not required for executing the learned skills
•Motor cortex, however, is essential for acquiring the subcortically generated skills
•This suggests that motor cortex “tutors” subcortical motor circuits during skill learning

[Deborah Zelinsky]

Very interesting summary, indeed! But, when I tried clicking on the link, I was told to buy a subscription. Don’t SFN members have this automatically?

[oliver_carter]

Agreed. Very interesting findings, indeed–and seemly so un-textbook-like that it would easily go viral. (Is that the right way to say that? I’m such a caveman in these new hip IT terms… hee, hee, hee…) I hope to follow up on that paper, since it will have some more things to look at ‘down under the hood.’ (referring to sub-consciousness state events)

Yeah, @mindeyeconnection (I guess I am encouraged to use that, rather than the actual name? is it? or is it by permission of the person, or what-have-you, that one can use a person’s real name? Not sure here, so, just for safety’s sake) , I do believe that even with a link from here, if the one linking does not have the journal subscription, a pay-per-paper (at least) choice is the only option. For me too, I would have to purchase at least the PDF usage.

For myself, I’m fine with that much, but would love to see a tad bit more information and ‘discussion’ of a sort, in the opening post of a thread. (as long as no copyright laws are overly bent)

Thanks for the share, @jkubie … OH ! It’s you !! I of course recall you from the old site, but could not guess it in my wildest dreams from just the username/avatar here… hee, hee, hee… Thanks !

[jkubie]

The paper and the videos are available at the Olveczky Lab web site:
http://www.oeb.harvard.edu/faculty/olveczky/

[Charlie_Wilson]

Interesting study. I wonder whether these findings can be thought of in the context of suggestions (more from human and some monkey studies) that the driving hierarchy in frontal cortex is a fronto-caudal one.

So this is the idea that increasing complexity (be it temporal or abstraction) is processed more frontally in frontal cortex, whilst more routine tasks, are processed more caudally. So this is the work of, for example, the Koechlin and Badre groups. These ideas centre on control of cognition, but one might generalize from the current findings that the nature of the control, even very motor control, is changing through learning.

Any thoughts?

[jkubie]

for paper and videos, try the Olvekczky lab site. http://www.oeb.harvard.edu/faculty/olveczky/

[jkubie]

The authors attribute the savings to “sub cortex”. I don’t think there is consideration of posterior cortex. I’m thinking basal ganglia.

[Charlie_Wilson]

I was mainly thinking about the role (and later non-role) of the cortex. In the sense that the “tutoring” function they describe of motor cortex might be able to be interpreted in the context of these ideas about (cognitive) control

[jkubie]

Interesting to speculate on a neural mechanism for “tutoring”. Presumably the learned pattern is due to changes in synaptic efficacy within the “sub-cortical” region. Let’s imagine that motor cortex projects to the circuits involved. I don’t think simple Hebbian logic can do this. Seems like an AND function, that Hebb logic can’t do. What are alternatives?