If you're a sentimentalist about ethics, and a constructionist about emotions, then it makes a lot of sense that a person's moral judgments would be different in another language.

I think the way this would work is something like: I witness an event that evokes an affective state based on my expectations and past experience with things I categorize as "similar." Then, I label my affective state with an emotion word. Then, based on that label, I engage in behaviour that I've learned is an appropriate response to that label. In this case, the behaviour would be making a moral judgment I guess.

It's true that the emotions and behavioural response are culturally learned, but also my internal concept of "angry" might be different from my concept of "fâché" simply due to the memories and associations that I have with the two words, even though they're supposed to mean roughly the same thing.

If this account is right, then someone is probably going to be inconsistent in their moral judgments depending on what language they're speaking (in addition to all the other environmental factors that are influencing them at the time).

So, yes I think you're right that it is about culture, but I don't think that makes it less about language!

In the article they say both native English speakers and native Spanish speakers had similar results in each other's language. Tbf though these are probably small studies, and culture might matter as well, and direct social environment you speak the language in, like French with your family and friends, English at work, and German with your SO's family.