In order to intervene effectively in this family, I needed to find out how the family operated as a system. I wanted to know what the natural sequence of interaction was. The best way to do that is to create a crisis, which is something that most family therapists avoid. If I make everything nice and lovely and warm, then I don't get down to the nitty–gritty. So I usually mention the most taboo things in the world for the family.

Virginia Satir taught me this. A lot of people think Virginia doesn't do it, because she does it in a nice tone of voice, but Virginia talks about everything that the family doesn't want to talk about. My style may be a little bit closer to Frank Farrelley's in the way I go about it, but it accomplishes the same thing.

So the family comes in and I say «Well, what are you doing here? What went wrong?» Immediately the mother says «This lousy kid over here has been getting out of hand.» I might turn around and say to the son «You son–of–a–bitch!» And then I ask the mother «What has he been doing, swearing?» Immediately the family goes into crazy land, and the system begins to operate. I can say «Well, what do you do if he does this? You probably don't scold him or anything.» She'll immediately start in quotes «Well, I tell him blah, blah, blah» and then immediately the kid will lose quotes and say «Look, goddamit, get off my back!» Then the father will say «Can I get a drink of water around here?» As soon as the family system starts operating, I sit back and observe, because I want to know how the family system operates without me. If it starts to slow down, then I step in and kick it to get it going again. I find out what the really sensitive areas are, so I can keep mentioning them to keep the family going.

This also wears them out, which is really useful. That's one of the things that makes my job easy. I've tried for a long time to train students to do this, but they get caught up in the content of what the family is doing, rather than stepping back and letting the family fight it out so that they can find out how the system operates.

The program in this particular family was really interesting. When the mother spoke, the husband responded like crazy. He went into what psychologists call «massive denial.» He climbed into the back of the chair and hid in the cushions. The oldest son was a carbon copy of the mother, and fought right back at her «RRrrrrhh!» And the more he fought back, the more the mother attacked him. If I interrupted the mother's behavior, the son kept attacking, but the father relaxed. That's important to know: the father was not responding to the son; he was only responding to what the mother did.

Woman: What did you do so that the father relaxed?

I shut the mother off for a while. When the fight got rolling, I just got up and stood in front of the mother, and the son yelled right through me. As soon as I cut off the mother visually, the father sighed and relaxed, even though the son was still screaming at the top of his lungs. When I stepped out of the way, the father immediately tensed up again. You can't do that kind of testing if you are glued to your chair the way many therapists are.

In this family, the younger son responded positively to his older brother. And when the mother went after the older brother, she might as well have gone after the younger one, because he responded as if the mother were going after him. He was a completely vicarious human being. If you talked to him directly, he always looked behind himself, no matter where he was sitting. He actually did that. I asked him «What do you think about this?» and he looked behind himself and said «Ah, well, ah … I don't know.» It was as if he weren't all there. But he really responded to whatever the mother did, even if the mother did it to the father or to his older brother.

The mother fought it out with me tooth and nail, and she was almost my caliber. She could hold her own against me, and there aren't too many people who can do that. But I have some really underhanded ways of fighting. I can switch logical levels so fast that I kept a little bit ahead of her, but I worked hard to do it. There were two male students and one female student in the room with me, and whenever the female student spoke to the mother, her behavior completely changed. The female student said things like «You are so unfair to your son.» The mother turned around and said gently «Well, now, dear, some day you are going to be a little older and you are going to be in my place… .«It was a completely different program. If a male had said that to her, she'd have boxed his ears off!

The mother's programs for communicating with men and women were totally different. The little girl did weird things in the session— things like getting up and knocking papers off the desk, interrupting, and making noise. If the son even took his eyes off what was going on, she'd shout «Pay attention!» But the little girl was safe from that.

Woman: And you didn't directly comment on that at all? You just watched it?

What good would it do to talk about it? If I tell them all the things I make distinctions about, that would make it easier for them to stay the same.

In order to test what I had observed, all I had to do was switch back and forth between acting like the son, the father, and the little girl, and see what different responses I got from the mother. I could actually get different responses from her by adopting the little girl's analogues. She began to respond to me in a way that mixed how she usually responded to men and to women.

There was just no way in the world to get the mother to attack the little girl. I asked «What's the worst thing the little girl's ever done?» She said in a sweet voice «Oh, one time she spilled blah, blah, blah.» When the mother talked to the little girl, the entire family loved it. They wished the two of them would run away together! They all responded positively, because the little girl got treated the way they all wanted to be treated. If the little girl communicated to the mother, the mother responded positively, but if the little girl communicated to one of the other people, the mother did not respond. That's very important. If she did, I could have made trickier interventions. I could have gotten the little girl and the brother going, and gotten the mother responding to that. But the mother didn't respond positively to anyone in the family except the little girl communicating directly to her. Everyone in the family responded to the mother.

So I had to figure out what I could get this little girl to do, to get the mother to respond in a way that would get the rest of the family to make the changes they wanted. When I first learned family therapy, I was told that everything works in triads—that when three people communicate, if person one communicates with person two, person three is always going to respond to that communication. It's not true. You can get them to respond to it, but they're not necessarily doing it already.

What I want to know in any family is what they are already doing, because then I can use what's going on now in order to change the system. This is a very important principle: How can I introduce a small change that will channel all the interactions in the family system in ways that force the system to change itself? When you can do that, the family system will do most of your work for you. If I want everybody to change in this particular family system, then I'm going to change the daughter. She will alter the mother's behavior, and ultimately everyone else in the system will change in response to the mother. However, it doesn't work the other way. If I had changed the younger son, it wouldn't have affected anyone else, because no one in the family responded to him. He was about as close to non–existence as it's possible to get. It was «to be or not to be» and he wasn't.