Dan Pink's insights about motivation, as found in the two videos below, are focused on how to improve business productivity, and they seem applicable to language learning as well.
His basic argument is that productivity goes down for activities requiring even a rudimentary amount of cognitive ability (e.g., language learning) when the primary motivators are extrinsic (e.g., a monetary bonus) as opposed to intrinsic. And this all appears to be backed by some 40 years of research.
So what are the intrinsic factors that increase productivity? Autonomy, mastery and purpose. As Dan's puts it, "autonomy is our desire to be self-directed" and "mastery is our urge to get better at stuff". Having a purpose means some kind of transcendent purpose our actions our working towards.
Applying this to language learning? Autonomy as a motivator seems to jive quite well with my inclination to avoid classes; very few classes could ever be called "autonomous". It also jives quite well with lots of people's advice to get exposure to the language however you want (e.g., through comics or movies, as opposed to through some rigid curriculum).
Mastery is a no-brainer when you're talking about language learning or simply learning in general; you should by the nature of the activity be getting better.
Applying Dan's idea of purpose might be a little harder. Grades, as an extrinsic motivator, would run afoul of Dan's ideas about motivation and I imagine that he wouldn't be surprised in that it doesn't lead to great productivity in language learning.
On the other hand, I look at what my own purpose has been in language learning, and it's primarily been economic; I learned the languages I learned because I thought they would be economically useful for me. Perhaps that's part of a more transcendent purpose of providing for my family or the like, or perhaps I was actually just motivated by wanting to communicate with people. The latter would make sense given that I've only ever gotten good at the languages in the countries I've been in.
The videos, after the jump.
Below are two videos of Dan laying out his arguments. They overlap a good bit, so you can just watch the shorter one (i.e., the top one) to get the gist, if you want.