One of them was from a talk by Bill Dodd, ‘Semantic prosody’ in FL teaching and learning’. He took the now-near-chestnut of the verb ’cause’ from English, and looked at three different ’cause’ verbs in German, finding them to have the same semantic bias.

[In a corpus of English, you'll find that the overwhelming majority of occurrences of 'cause' as a verb have collocations that are bad things: viruses, negligence, vandals, infections, assaults, and deficiencies cause devastation, delay, injury, nuisance, consternation, havoc, death, ruckus and loss; not a lot of kittens causing giggles.]

Dodd looked at verursachen, bewirken, and hervorrufen in the corpus of the Digitale Wörterbuch der deutschen Sprache and found (as you may see yourself, either in deutsch or in google’s translation) that all three had exclusively nasty neighbors: Schade, Kosten, Brand, Störung, Verlust, Geräusch, Tod, Kopfschmerz.

I haven’t looked at evidence from other languages, but posit for a moment that all languages’ verbs of causality are associated with negative things. What does it tell us about causality, and the circumstances in which people think about causality?

Maybe people simply don’t often stop to ask why good things happen. Good things are exactly what are supposed to happen all the goddamn time; good things are what I deserve and there’s no “cause” for them: they are the natural order of things. So maybe a person looks for a cause only when they’re disappointed by an effect.