nd I wouldn't necessarily object to it if British society were largely organised around tribal units.

I think the real problem is that we lack information on other types of family structures - ie. we know when marriages are formed and break up. We don't know how many stable opposite-sex partnerships there are/how long they last on average. I guess we'll get some data on same-sex couples after December - but then it'll only be the ones who choose to register their partnership. We don't have detailed information on single parents (ie. in some of these cases the parent will have a partner they see occasionally, in others there'll be a live in grandmother, or they'll have a support network of friends/family around them. And this is without entering the entire realm of polyamorous relationships. Marriage is only seen as more stable because it's the one we've got lots of data for. I wonder if there's a clever DPhil student about who'd want to start doing research into this to find out if marriage really is a more stable family structure.