So there's a very widely used model of mood that posits this (the author didn't cite this which was surprising to me). The PANAS (https://en.wikipedia.org/wiki

So there's a very widely used model of mood that posits this (the author didn't cite this which was surprising to me). The PANAS (https://en.wikipedia.org/wiki/Positive_and_Negative_Affect_Schedule https://en.wikipedia.org/wiki/Positive_and_Negative_Affect_S...) is a widely used measure of mood that's based on this model, as an example. One axis is "negative mood" and the other is "positive mood", so it's closer to something like despair on on one end of one axis, and joy on the end of the other axis, and lack thereof on both ends of each axes. So high negative emotion and low positive emotion is something like sadness or something; high negative emotion and high positive emotion might be like anxiety. There's arguments about which model is right (another is negative versus positive mood on one axis [i.e., valence], and arousal versus lack of arousal on the other axis). The issue is that depending on time frame, or what you're assessing, people can feel positive and negative emotion at the same time. But on micro timescales I think there's more arguments. It might just be a sort of descriptive thing without any "correct" answer. When you ask people to rate their mood, you find two clusters of responses, roughly corresponding to negative and positive mood axes, but they're slightly negatively correlated, suggesting one overarching valence dimension, but not as dominant as you might intuit.