Drugs can be classed into two categories. Those we know the mechanism of action and those we just know what effects it produces. A lot of mental health meds fall into the latter category.
There's a third class: drugs that were discovered because they caused an effect that was theorized to be related to the actual target by long-disproven ideas of how bodies worked, and that show little or no benefit to the target condition but are still prescribed as they have always been.
We decided that seizures made people sane, created drugs and treatments to spark controlled seizures, then tried to make drugs that seemed similar but didn't cause tardive dyskinesia. Tardive dyskinesia was the only reason these drugs were ever even investigated, but the field somehow still moves forward like a racehorse without a head, claiming effects 10% above placebo.