On 'patriarchy', I agree with you that a situation with patterns that favor men over women tends to get ideological explanation that claim this is a 'natural' state of affairs, which leads to discrimination (i.e. judging a person as intelligent, or competent, or ... simply by his/her sex). I'm not sure if 'discrimination' and 'misogyny' are the same thing; as I understand it, you could be 'misogynous' (meaning you don't want to be around women -- like a misanthrope who doesn't want to live around people and prefers to be alone) without discriminating against women in any way other than avoiding their company. But I do agree that misogynous people would tend to discriminate against women and spread bad female stereotypes because they would rationalize their preference by saying women are indeed bad, stupid, incompetent (people apparently always want to offer some rationalization for their preferences...).

But I'm curious about the definition you give -- about 'patriarchy' being (the situation that obtains when) the man is the head of the household with women and children (legally? customarily? physically?) dependent on him. Is the mere pattern sufficient, or does it have to be abused -- by the creation of anti-female stereotypes to justify the pattern -- before you can say there is 'patriarchy'?

Let me give you an example of a sexual segretation pattern that, as far as I know, has never been claimed to be misogyneous (or mysandrious = anti-male either): naming. Certain names are only given to women (Mary, Jean, Laura...) and others only to men (Martin, Robert, Stephen...). As far as I know, this pattern -- the 'gender character' of names -- has never been ideologically used to claim superiority of one gender over another; so its existence has never led to, or been claimed to imply or to explain, 'discrimination'.

You had said that physical abuse only should be usable as a defense. But what about cases like the one I mentioned in my previous post above? I also thought of cases like the one described on this page: http://www.wendymcelroy.com/dvblog/dv004.html . The only 'documentation' you could get of the abuse pattern was when she stabbed her husband on the thigh -- and since he claimed that was nothing, I'm not sure that could be used as proof of the existence of a pattern. Most of the abuse was of the kind that is not easy to document (like turning the children against him), unless he started audio- or video-recording her abuse sessions. In other words: I can imagine cases of real long (non-physical) abuse that could include threats to the life of the spouse and/or children, and a true fear for one's (or one's children's) security, and yet would be hard to document and prove on any way other than the testimony of the defendant. What do you think?