It's hard to avoid those harrowing images of those thousands of people fleeing war and destruction. The pictures are everywhere and the effects of these massive movements of people are wide and far-reaching. Post-World War Two, many psychologists turned their research to examining how we establish differences between 'Us' and 'Them'. Turns out it is disturbingly easy to form what are called 'in groups' and 'out groups', often on the most arbitrary of criteria, and with lasting consequences. (When you read some of the research, it's tempting to conclude that you, personally, would never do some of these things that dehumanise others, but the uncomfortable truth is that we are all prone to this kind of thinking).
The September issue (2015, Vol 28 No 9) of The Psychologist carried two articles on current research into how and why we dehumanise others. Often we do it because we are afraid, and the latest research suggests that in times when we feel particularly under threat our propensity to de-humanise is increased. The extent to which we hold an 'out-group' as less-than-human acts as a predictor of our other attitudes. For example, there was more dehumanising of Arabs by US citizens immediately following the Boston Marathon bombing, which showed up in how this group, in general, should be treated as a consequence.
Philip Zimbardo (he, of the famous Stanford Prison Experiments) argues that dehumanisation is a form of evil and that we are all capable of it if the situational context is supportive of such. The process of dehumanisation diffuses responsibility - it allows us to take actions that would otherwise be highly uncomfortable, and possibly do things to others that we would never dream of doing in other circumstances.
However, he also pointed to something that he called 'heroism' - something that requires the courage to act in a different way. This can be learned and he has established the Heroic Imagination Project to teach school children how to challenge mindsets, question peer group pressure and overcome stereotype threat.