Secret computations of the hidden brain 3: Dopamine and learning
by Brandy Eggan and Jim Stellar
This post is the third and final one of our “Hidden Brain Series” that focuses on the role of dopamine in the brain’s reward circuit. Thus far, we hope we have shown the importance of this neurotransmitter and how its primary brain circuit appears to have a sort of intelligence on its own, constantly adapting its release patterns in response to reward history and learning. Moreover, dopamine-related brain circuits also have the ability to drive the execution of motor behavior surrounding a rewarding event, a very important function as what matters in nature is outcome. To complete this series on hidden brain computations, we finally touch on the role of striatal dopamine neurons in different areas in executing active learning versus producing old habits.
Yet another dissection of the striatum
Habitual behavior is a previously learned part a routine. Placing your sneakers on your feet and your ability to tie the laces while your mind is focused something else, like the route you are soon planning to run, is an example of a habit. No one thinks about the laces.
Active learning is different than habitual behavior because it requires a conscious combination of movements and observations of the consequences of those movements that leads to goal-directed behavior. Say when you are running around the block and you come across construction that has broken up the side walk. While you carefully navigate and attempt to not lose balance the first time, your brain registers this event. With a few more runs through this area, you may find yourself remembering which stones are loose and without thought you begin to avoid them. This is when your active learning begins to switch to habitual behavior as a result of repeated exposure.
The striatum (see the diagram above that is taken from blog post 2 of this series) contains different neurotransmitter/receptor populations and different neuron shapes/types that allow it to be divided into regions. One such division is a differentiation between the dorsomedial and dorsolateral portions of the striatum.
In this division, the dorsomedial striatum (DMS) is thought to be critical for active learning. Research has shown that food deprived animals with a deactivated dorsomedial striatum have a significant impairment in learning to lever press for a food reward. Moreover, if the experimenter shifts the criterion and requires more lever presses for the same reward, it is highly unlikely the animal will be able to actively learn the new paradigm at all (Yin et al, 2005).
Our brains do this kind of learning and re-learning all of the time. Think back to the laces example, but when you were a child learning to tie your shoes. The association here is that tied shoes result in a decreased likelihood of tripping, getting hurt, and maybe create a certain pride that you could act like an adult by being able to tie them yourself. A child with impairment in their dorsomedial striatum would struggle to learn this task, regardless of the number of times they find themselves face down on the sidewalk after having tripped. After a long time (in comparison to other children the same age), the child may eventually learn the task, but if the parents suggested changing the knot style to a double knot to prevent the shoes from coming untied, it is again unlikely that the child with DMS impairment will be able to learn this variation on the task.
On the other hand, the dorsolateral striatum (DLS) is believed to play an important role in the execution of habits that are already learned. Habit execution also can occur outside of consciousness, thus separating them from the learning in goal-directed behaviors where one must concentrate on what one wants to achieve. Research has shown that animals can be trained to bar press for a food reward, which develops in to habit. In this study, habitual behaivor can be differentiated from goal-directed behavior because if the food reward is made unpalatable, the animal will still press the bar when placed in the testing chamber, they simply will not consume the devalued reward once it is acquired. That habit of bar pressing is still present, but a food reward (previous driver of goal-directed behavior) is now not necessary to motivate the action. Interestingly, if an animal is trained in the same bar pressing paradigm for reward and then the reward is devalued (sweet treat made bitter), if their dorsolateral striatum is deactivated, they will not resume the bar pressing habit when placed back in the testing chamber (Yin et al, 2004). The studies controlled for duration and type of experience so it is not the case that the rats in the first group simply had less practice than rats in the second group.
Our brains use habit every day – often reward is reduced or even ceased after a period of time. No longer are we tying our shoes out of fear of tripping, as a matter of fact you probably do not even remember the last time you have fallen due to untied shoe laces, and we are sure you don’t remember the last time your mother praised you with a “good job” after you had executed the perfect crossing of your “bunny ears” with the folded laces. Rather, you are now tying your shoes because that’s just what’s engrained. These processes allow you utilize the highly efficient unconscious of the brain to execute specific learned actions while your conscious permits for active learning as well as planning out the next steps in your day.
Application to applied learning experiences
While the dorsomedial and dorsolateral striatum sound suspiciously like the two components in Kahaneman’s famous 2011 book Thinking Fast and Slow, we are not making a claim that this is the mechanism – only that such a separation exists. Since this blog has often used Khaneman’s book as a jumping off point for internship, let’s shift gears and look at what college students learn on an internship. Let’s say that they join an office in a specific profession and work there for months. They get confortable with the office routine and with the specific knowledge of that field, say a law firm or a clinic. They learn skills, see clients, and end up feeling welcome rather than as the “new person.” Moreover, soon enough their daily work seems to be getting easier. Why is that? We would say that it is because the tasks they found themselves actively learning over the first few weeks are becoming engrained to habit and shifting being controlled by the DLS to the DMS. That is, they no longer require conscious attention. In the beginning of any new position, students feel like a “deer in the headlights” when presented with a task, but after a while they become “old hands” with the daily, routine. We could say in this way they have acquired expert knowledge. The expert knowledge allows the student to have the cognitive resources to focus more on new stuff, perhaps on the client and their specific needs rather than just selecting and applying the relevant knowledge. It could permit the student to do deeper learning and observe more about the law or medicine in the firm or clinic using again the dorsomedial striatum to do that deeper learning. After all, the brain is wired to run in the most effective fashion in regards to information processing, storage, and recall.
All of this is speculative. What is not speculative is that none of us are generally aware of how we learn or that we have the advantages of a brain with operational regions that are largely hidden from our conscious awareness and that do some of this learning for us. We just sort of ride on top of the process and take the intellectual lessons from the learning, like whether we truly like the profession we are experiencing. The fact that we do not know how we do our own learning emphasizes to us as designers of the higher education experience, the need to have some learning take place in an authentic, substantial, real-world environment where the profession is practiced. Someday, when we know how we learn and when we know the professional practices to be learned, we may be able to do a powerful simulation or just have a text-based course that teaches us everything we need to know from the direct experience. But for now we must use the tools available to us. We must push our students into applied learning experiences, that internships, volunteer projects, community service programs, etc. to develop of those practical skills in applying knowedge. It is here, during these experiences, that we believe students will develop a unique skill set, or expert knowledge that one can just not get from the traditional classroom and that is going to be the key to student success once they leave college.
Kahneman, D. 2011. Thinking Fast and Slow. Farrer Strauss & Giroux. New York, NY.
Macpherson, T., Morita, M., Hikida, T. 2014. Striatal direct and indirect pathways control decision-making behavior. Front Psychol. 5: 1-7.
Yin, H.H., Ostlund, S.B., Knowlton, B.J., Ballenine, B.W. 2005. The role of the dorsomedial striatum in instrumental conditioning. Eur J Neurosci. 22: 513-523.
Yin, H.H., Knowlton, B.J., Balleine, B.W. 2004. Lesions of the dorsolateral striatum preserve outcome expectancy but disrupt habit formation in instrumental learning. Eur J Neurosci. 19: 181-189.