That still sounds like just one dimension to me. For two dimensions, you would need "mild very positive emotions" (contentment?) and "intense slightly negative emotions" (overpowering nostalgia?).

One way to get around this is to classify emotions into active and passive (or high- and low- arousal), where, for example, anger would be active/negative, and grief would be passive/negative. Like the emotion diagram I've seen around the internet lately:

Emotion Diagram

That being said, it's still interesting how vague emotional classifications can be.

That still sounds like just one dimension to me. For two dimensions, you would need "mild very positive emotions" (contentment?) and "intense slightly negative emotions" (overpowering nostalgia?).

The two dimensions are negative and positive: you can be both negative and positive at the same time, so your degrees of negativity and positivity can be treated as a point in a two dimensional space.

This still isn't a fully two-dimensional space -- or at least it's not a square, since extremes of either positive or negative arousal tend to suppress the other. WIthin a certain range, though, you can be both negative and positive or neither, as well as one or the other. So negative-positive isn't directions on an axis, it's a pair of sometimes but not always anti-correlated measurements.

Maybe we can apply the virtue of scholarship in a differentiated fashion depending on the field.

For philosophy, psychology, and anything "harder" than that: do scholarship, they frequently experiment or make rigorous arguments.

For sociology: spin your own theory, that's all sociologists are doing anyway and your theory doesn't have to sound impressive so you can get tenure.

Sounds like a plan!