Our ethical decisions may depend on the language they are couched in

Here's an interesting thing. You've probably heard of what's usually referred to as the "trolley problem" of ethics and philosophy. So, people have to decide whether to sacrifice one person in order to save five others who are about to be killed by a hypothetical runaway trolley bus (or tram, or train). There are various formulations and variations, like what if it was a fat person or an old person, or what if you had to actually push a person off a bridge in order to save the five rather than just pull a track switch. 

Generally speaking, a sizeable majority think that it is morally permissible to make the utilitarian choice and pull the lever to save the five at the cost of the one. However, much fewer people would push someone off the bridge to save the five, reflecting the increased "emotional resonance" of such an action. 

More recent studies have looked at what difference the language in which the problem is posed makes to people's moral decison-making. The problem was put to people who spoke more than one language, both in their native language and in their second language. By a substantial margin, more people chose to pull the switch to kill the one person inorder to save the five if the problem was put to them in their second language, rather than in their native language. The difference was even starker for the more emotive problem of having to push someone off a bridge rather than just pull a rail switch at a distance. Furthermore, the poorer people's mastery of the second language, the more marked the effect.

This suggests that a native language holds much more emotional resonance, while a second language maintains more psychological distance for most people. This makes some intuitive sense, I guess, but it's interesting to see it demonstrated in quite such a stark manner.