[curi]

So suppose we are studying sad people, and we randomly give some of them a chemical, in a controlled and double blind way. And those people become happy, and the others don't.

Then, it could be that the chemical somehow causes happiness. Or it could be that the chemical causes memories of one's children, and everyone in the study has happy memories of their children.

So, the causation is still in doubt. Many explanations are consistent with the observed data.

---

Yes I am proposing that chemicals correlate with "aggression" in humans and rats for different reasons. But I don't accept that the thing called "aggression" is the same thing in humans, and in rats. To help illustrate why not, consider aggressive behavior by video game characters. Same word, but not the same phenomenon -- in that case, no emotion is present, and the cause is some C code.