How does muscle memory work?
From smiling to playing musical instruments, our brains have a tendency to take certain physical skills and keep them on file almost permanently.
I got a mandolin for Christmas this year (the eight-string miniature guitar-ish thing, not the finger-mangling kitchen utensil, though they kind of have that in common), which has been so exciting to learn since it’s slightly different from my longtime musical companion, the guitar. What’s really struck me while I’ve been noodling around on it with the new scale and chord patterns has been the ease with which I can switch back to the old faithful six-string. It feels like my hands just know where to go and what to do, despite all my musical brainpower being diverted to the new concepts.
At some point I expected to rattle off some atonal, earache-inducing sounds on the guitar after a mando session since those shapes and patterns were front of mind, but it hasn’t happened. (The outcome of this week’s mando practice is a half decent rendition of Emerson Lake and Palmer’s I Believe in Father Christmas, for the record.) The part of my brain that locked down the guitar, that helped my mom hop on a bike this summer after many moons, that sees people spin through Rubik’s Cube algorithms with their eyes closed, and that might lead you to remember the steps to a dance you learned in high school P.E., is obviously doing some serious work behind the scenes.
This is a video from 2019 of Marta C. Gonzalez, a prima ballerina who danced with the New York Ballet in the 60s. She passed away last year, but at the time had been diagnosed with Alzheimer’s and was living in a long-term care facility in Valencia, Spain. Her healthcare aide played her the music from Swan Lake and, despite having lost the ability to speak and to recognize those around her, she broke into the choreography she learned over 50 years ago.
The body and mind are capable of teaming up for some serious long-term memory surprises that seem to happen unconsciously. Without having to think about it, and sometimes only encouraged by context or a sensory trigger, the body intrinsically knows what to do in these situations. The lowest common denominator in many, if not most, of these situations is the repetition with which the skill was learned.
Repetitive practice under consistent circumstances has been found to help Alzheimer’s patients (re)gain motor skills. Similarly, young children are found to learn specific motor skills better under semi self-guided conditions, rather than totally guided conditions. The kids who figure out pattern chains (like hopscotch or a clapping game) on their own tend to learn faster and remember better (and rely on conscious self-instruction less) than those who only learn through being instructed by someone else. The connection between consistent learning conditions (environment) and the level of self-instruction (independent practice) are two key components to establishing muscle memory for gross motor skills.
So what’s the mechanism for this phenomenon? The box of ultra long-term physical muscle memory is stored somewhere in the brain, but what’s the trick to unlocking it? How does the brain decide what’s worth storing there? Do the other areas of the brain need to work to access it?
I have lots of questions.
Let’s talk fundamentals first. For a long time, researchers thought muscle memory was entirely learned. Humans appeared to be blank slates who would take on common gross and fine motor skills as we hit certain developmental and social milestones specific to our environment. Think running, writing, and facial expressions. Eventually, researchers observed skills they thought were contingent on observation in people born without eyesight. Since people who are blind use smiling, frowning, and other facial expressions in much the same ways as people born with sight, there’s evidence that at least some muscle memory is hard-wired into our genetic code.
The second big piece here is the two types of memory we’re talking about. Not necessarily short-term and long-term which deal with storage time, though the distinction is relevant, but declarative and procedural which deal with access - the concept of conscious vs unconscious recollection.
Declarative memory refers to all the stuff you need for a championship run on Jeopardy!, RIP Alex Trebek. The facts, figures, people, places, and things that you retrieve when you consciously retrieve specific information is filed under declarative memory. Procedural memory, on the other hand, is the stuff that’s hard to put into words. It allows you to undertake behaviours like your shower routine, driving a stick shift, and riding a bike, all without paying too much attention to what exactly your limbs are doing to achieve the thing.
Good? Good. Now let’s get to the connection between motor skills and the areas of the brain that make them happen. From our friends at Wikipedia:
When first learning a motor task, movement is often slow, stiff and easily disrupted without attention. With practice, execution of motor task becomes smoother, there is a decrease in limb stiffness, and muscle activity necessary to the task is performed without conscious effort.
This is where it gets good. In order to reduce the amount of conscious focus directed to the limbs executing the movement, there has to be a shift in how the memory of doing the movement is encoded.
The neural pathways that are key players in muscle memory encoding, specifically, are the prefrontal and frontal cortexes and the cerebellum while someone learns a skill. These regions are responsible for executive function, reasoning, attention, coordination, and timing, but generally aren’t really helpful in the storage of skills. The prefrontal and frontal cortexes along with other motor centres in the brain actually become less active once the skills are learned.
The learning/habit-forming, declarative memory processes can be linked more to the basal ganglia, which have lots of stimulus-response neural connections with the cerebellum that grow stronger over time as new motor tasks enter the repertoire. This structure is central to motor movements, habit-forming abilities, and emotion.
If I understand correctly, this is where the connection is between contextual-repetitive learning and long-term recall happens. This is why Marta C. Gonzalez could retrieve choreography but couldn’t communicate using language - the neural connections between hearing Swan Lake (the stimuli and the context) and the muscle memory of her performances (the response and the repetition) were strong and didn’t rely on her other brain regions, beset by Alzheimer’s, to put the pieces together. She didn’t need any help from prefrontal or frontal cortexes, nor did she need to consciously recall each movement as it was taught to her. The procedural memory and basal ganglia-cerebellum connections did the work below the surface for her.
If you’re less of an amateur than I am about neuroscience (hello, two undergrad courses once upon a time), please use the comments to correct my interpretation of things I learned on the internet.
The next thing I want to explore is how exactly the brain locks down short-term memories for the long haul. In 2010, Dr. Liangsuo Ma and with a bunch of colleagues studied the role of motor learning on activity levels in different regions of the brain and found that:
Muscle memory consolidation involves the continuous evolution of neural processes after practicing a task has stopped. The exact mechanism of motor memory consolidation within the brain is controversial. However, most theories assume that there is a general redistribution of information across the brain from encoding to consolidation. Hebb's rule states that "synaptic connectivity changes as a function of repetitive firing." In this case, that would mean that the high amount of stimulation coming from practicing a movement would cause the repetition of firing in certain motor networks, presumably leading to an increase in the efficiency of exciting these motor networks over time.
Basically, the brain registers the consistent, repetitive stimulation while we learn a new skill or form a new habit and that’s the cue it needs to transfer that muscle memory over to long-term, procedural memory stores. It’s that easy.
Everybody’s heard of the adage that it takes ten thousand hours to become an expert in something. The catch here is that the repetitive nature of building muscle memory isn’t necessarily a guarantee of expertise. Take piano, for example. If an aspiring pianist practiced their scales and took on increasingly challenging pieces to the tune (sorry) of ten thousand hours, but did so clumsily and with poor timing or technique, they’ve just encoded muscle memory that’s going to be increasingly hard to change if they want to improve. Muscle memory is only as good as the repetition itself, and habits become harder to change the longer they remain encoded.
Muscle memory works because of the actions themselves, repeated over time under similar circumstances, that build strong neural pathways outside of the conscious information/language centres. These memories (more like habits by some point) are rooted in the body’s motor function, a more innate and fundamental physical expression of our person-ness.
There’s a great quote from a baseball player that I’m about to butcher, but it goes something like, “the best thing you can do when you step into the batter’s box is forget everything you’ve learned about baseball.”
The muscle memory will do the work; thinking just gets in the way.