At the moment I am observing one of the most mundane things with my horses, and still I find it ever so fascinating. The situation is that I let them out into another pasture for half an hour in the mornings and evenings. Of course they are very fond of the grass, so when I have to take them back into the winter pasture, that's not exactly what they would choose to do all by themselves. I walk up to them, ask them to lift their head from the grass, and then we walk back in together. I either just walk next to them or lead them by holding their mane.
It happens every day, so they are learning that when I come to them after a certain time, it always means leaving the grass – which they'd rather not do. This year I have not used any "special reward" waiting for them in the paddock (just a few grains on the way, but that is less valuable than fresh grass). Therefore, there is no positive consequence of coming with me. Neither do I use pressure to make them come or threaten to chase them away if they don't. So the process of me leading them back in is about as neutral as it can get, and being inside is the worse option. Accordingly, I was expecting that over a few days their willingness to come with me would greatly diminish, and that perhaps they would even start running away from me in this situation.
However, the opposite has happened. Walking up to them still is as easy as it was in the beginning, and walking away from the delicious pasture together is getting easier
: while leading them by their manes was no problem at any point, now they often move towards the gate all by themselves, or sometimes even trot there so that I can just let go of them and follow them while they go back in all by themselves.
The reason why I think it's mundane is that everyone knows about the power of habit. The things you do every day become so easy that you don't even have to think when doing them, and if you form an association between a certain stimulus (in that case: me walking up to them) and an action (walking back into the winter pasture), the mere sight of that stimulus automatically primes the representation of that action and thus facilitates or even triggers it. But I think it's so amazing that this seems to work even against a strong emotional factor (the effect of eating versus not eating the delicious grass). Therefore, I am wondering what positive emotional effects (in terms of the brain's reward system) it may have to perform a habitual action. I know that the anticipation or experience of difficult or conflicting actions can work as an aversive signal and trigger the corresponding brain processes, but I guess I will have to do some research on the emotional changes that go along with performing an action repeatedly so that it becomes a habit.
So I couldn't wait and have done some searching right now. Here is a little summary: The development of habits typically goes along with a reduced sensitivity for behavioural consequences (rewards) or their value. That is, once a behaviour has been trained extensively, it doesn't really matter anymore whether you are interested in the reward, you just do it. Applied to my pasture example, that would mean that once my horses got used to walking back in when I come, it doesn't matter that much anymore what effect this has for them (i.e. the lack of grass becomes less important). Now the question is why that happens.
One mechanism could be that during training, the link between the stimulus and the action (e.g. Romy comes and walking back in) is increased, resulting in a more direct and automatic control of actions by the environment, while at the same time the strength
of the association between an action and its outcome (e.g. walking back in and grass) is diminished. Such a central role of the association strength between action and outcome would be in line with research showing that habits develop more readily when the correlation between a behaviour and its outcome is low.
Interestingly, that does not seem to be the case, though. Instead, it seems that not the strength of action-outcome associations but their influence
on behaviour gets modulated, by an active suppression mechanism. That is, the associations stay intact but just cannot exert control over behaviour anymore. This inhibition of goal-directed responses allows habit-based behaviour to play a larger role in action control. However, it's not irreversible: Once that inhibition is removed (by temporally deactivating the brain structures responsible for it), the behaviour becomes goal-directed again, so that rewards and their quality regain their importance.