Nicole Rust on Nostr: Thrilling developments toward understanding mood Of all of the brain's functions, ...
Thrilling developments toward understanding mood
Of all of the brain's functions, mood is among the most mysterious. Mood is an ever-present feeling that we continuously experience, characterized by states like: happiness, sadness, anxiousness and calmness. It typically fluctuates slowly, subject to a multitude of disparate forces that we cannot always pinpoint; somedays we just wake up in a better (or worse) mood and we are not sure why. Mood is often differentiated from acute emotions that are more often targeted at something; for instance, we are afraid of tigers and we are disgusted by feces but more often than not, mood isn't directed at a specific thing. The general idea that's been floating around for awhile now is that mood is something akin to a running average about our overall well-being that exists to help motivate us to make good choices.
A groundbreaking paper in 2014 led by Robb Rutledge (now at Yale) demonstrated that this idea about mood isn't quite right insofar as mood tracks not with overall goodness (rewards) but rather unexpected goodness (reward prediction errors). In the context of a gambling task in which individuals had to choose between certain rewards and gambles, happiness tracked with unexpected wins (averaged over about 10 trials):
https://doi.org/10.1073/pnas.1407535111
That finding has been extended to the notion that mood reflects a sort of momentum that informs us about when things (like our environment) are changing in ways that make rewards more or less available. And the purpose of mood is to make learning efficient. Under stable conditions, reward-based (reinforcement) learning is a great way to learn, for example, what trees have fruit if you are a monkey searching. However, it's an inefficient way to learn if it's spring or fall and the overall amount of fruit is changing (in that case, you don't want to learn every tree individually). Instead, it makes sense to update your expectations about those changing conditions; this is what mood is thought to be for:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4703769
What's most exciting to me is the formalization of this tremendously important but very slippery thing we call mood into mathematical models that can be tested. This article does a beautiful job describing how that type of approach can be used to formalize: happiness as an emotion versus a mood; the difference between happiness and pleasure; and the difference between depression versus anxiety.
https://doi.org/10.1016/j.cobeha.2021.01.007
Of all of the brain's functions, mood is among the most mysterious. Mood is an ever-present feeling that we continuously experience, characterized by states like: happiness, sadness, anxiousness and calmness. It typically fluctuates slowly, subject to a multitude of disparate forces that we cannot always pinpoint; somedays we just wake up in a better (or worse) mood and we are not sure why. Mood is often differentiated from acute emotions that are more often targeted at something; for instance, we are afraid of tigers and we are disgusted by feces but more often than not, mood isn't directed at a specific thing. The general idea that's been floating around for awhile now is that mood is something akin to a running average about our overall well-being that exists to help motivate us to make good choices.
A groundbreaking paper in 2014 led by Robb Rutledge (now at Yale) demonstrated that this idea about mood isn't quite right insofar as mood tracks not with overall goodness (rewards) but rather unexpected goodness (reward prediction errors). In the context of a gambling task in which individuals had to choose between certain rewards and gambles, happiness tracked with unexpected wins (averaged over about 10 trials):
https://doi.org/10.1073/pnas.1407535111
That finding has been extended to the notion that mood reflects a sort of momentum that informs us about when things (like our environment) are changing in ways that make rewards more or less available. And the purpose of mood is to make learning efficient. Under stable conditions, reward-based (reinforcement) learning is a great way to learn, for example, what trees have fruit if you are a monkey searching. However, it's an inefficient way to learn if it's spring or fall and the overall amount of fruit is changing (in that case, you don't want to learn every tree individually). Instead, it makes sense to update your expectations about those changing conditions; this is what mood is thought to be for:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4703769
What's most exciting to me is the formalization of this tremendously important but very slippery thing we call mood into mathematical models that can be tested. This article does a beautiful job describing how that type of approach can be used to formalize: happiness as an emotion versus a mood; the difference between happiness and pleasure; and the difference between depression versus anxiety.
https://doi.org/10.1016/j.cobeha.2021.01.007