I feel as you do, but it's worth pointing out that if you come from a different frame of reference you may not be observing what you think you are observing. I can't say for sure in this case of course (hence how I ended my post), but he may have seen, say, discussion (in a language he may not be fluent in), showing each other papers, etc., yet may not have been seeing the actual 'cloning' of homework answers. If you walked into a bunch of American students doing this (but couldn't, for whatever reason), it's more natural to assume that there is direct cloning involved.

But different cultures have different norms.

Let's say, for example, that problems 3, 4, and 7 are assigned. Now, problem 7 is a bitch, but shares some of the devilish details, perhaps the 'trick,' with problem 6. Let's also assume that the course requirements are "You must do your own work on homework." As a team, the group worked problem 6 together. But one of the group was still struggling to complete problem 7 for homework. He turns to his friend for help, and the friend replies "remember what we did in problem 6? You do the same thing, but with two <whatevers> instead of one."

Is this cheating?

The direct help "with two <whatevers>" is getting close, but if you are from another culture you may, honestly, think no. Because remember, your whole understanding of education is about getting/doing whatever you need to do in order to solve the problem. So you are going to judge rules differently than someone whose first instinct is "this has to be my work only-- I'll fail all by myself." (I hypothesize that this is part of the problem with international IP laws as well.)

On a personal note, one the other side of the desk I was delighted when students teamed up on homework and did extra. In fact, making solution manuals available and publishing old exams was simply a sneaky way for me to get them to do more work.