I've recently been getting a lot out of the psychotherapy model of Internal Family Systems, as described in this book. I just wrote a comment on Slate Star Codex describing some of its basics and what I've gotten out of it, and thought that I might as well repost it here:

--

I recommend this book, though with the note that I often don't need to follow the full process outlined there. Sometimes it's definitely necessary, but what I've found even more commonly useful is something that it discusses at the beginning of the book, which it calls "getting into self".

Here's the basic idea. Suppose that a part of your mind is really angry at someone, and telling a story (which might not be true) about how that person is a horrible person with no redeeming qualities. Internal Family Systems says that there are three modes in which you might react to that part:

First, you may be entirely blended with it (for those familiar, this corresponds to what Acceptance and Commitment Therapy calls cognitive fusion). This means that you are experiencing everything in terms of the story that it is telling, and have forgotten that this is an emotional reaction. So you feel that it's just objectively true that the other person is horrible and with no redeeming qualities.

Or you might be partially blended with it. In this case, you realize that you are experiencing an emotional reaction, and that your thoughts and feelings might not be entirely justified, but you still feel them and might not be able to stop yourself from behaving according to them anyway.

Finally, you might be "in Self", meaning entirely unblended. Here you are still aware of the emotions and thoughts, but your subjective experience is that they're not your emotions, they're someone else's - they're coming from a part of your mind which is experienced as separate from "you". In this mode, you do not feel threatened or overwhelmed by them, and you can maintain a state of open curiosity towards whether or not they are actually true.

My experience is that usually if I have an unpleasant emotion, I will try to do one of two things: either reject it entirely and push it out of my mind, or buy into the story that it's telling and act accordingly. Once I learned the techniques for getting into Self, I got the ability to sort of... just hang out with the emotion, neither believing it to be absolutely true nor needing to show it to be false. And then if I e.g. had feelings of social anxiety, I could keep those feelings around and go into a social situation anyway, making a kind of mental move that I might describe as "yes, it's possible that these people all secretly hate me; I'm going to accept that as a possibility without trying to add any caveats, but also without doing anything else than accepting its possibility".

The consequence has been that this seems to make the parts of my mind with beliefs like "doing this perfectly innocuous thing will make other people upset" actually update their beliefs. I do the thing, the parts with this belief get to hang around and observe what happens, notice that nobody seems upset at me, and then they are somewhat less likely to bring up similar concerns in the future.

In terms of global workspace theory, my model here is that there's a part of the mind that's bringing up a concern that should be taken into account in decision-making. The concern may or may not be justified, so the correct thing to do is to consider its possibility, but not necessarily give it too much weight. Going into Self and letting the message stay in consciousness this way seems to make it available for decision-making, and often the module that's bringing it up is happy to just have its message received and evaluated; you don't have to do anything more than that, if it's just holding it up as a tentative consideration to be evaluated.

The book has a few different techniques that you can use for getting into Self. One that I often use is to try to get a sense of where in my body the emotional sensations are coming from, and then let my mind create a visualization based on those. Once I have a visualization and a physical location of the part, it's easier to experience it as "not me". Another thing that I do is to just make that mental move that I described - "okay, this is a possibility, so I'm just going to test it out". I find it useful to first stay blended with the part for a while, to get a sense of what exactly is the story that it's trying to tell, before unblending and getting into Self.

E.g. a while back I was having a sense of loneliness as I laid down for a nap. I stepped into the part's perspective to experience it for a while, then unblended; now I felt it as a black ice hockey puck levitating around my lower back. I didn't really do anything other than let it be there, and maintained a connection with it. Gradually it started generating a pleasant warmth, and then the visualization transformed into a happy napping cartoon fox, curled up inside a fireball that it was using as its blanket. And then I was no longer feeling lonely.

That said, sometimes a part is not content to just raise a tentative possibility; sometimes it feels like something is an emergency, so you must act right away. Obviously, sometimes you really are in an emergency, so this is justified! But often times it's based on the part having an unrealistic fear, which in the IFS model tends to be a result of some past trauma which it is reliving, not realizing that the circumstances of your life have changed and you're now capable of dealing with it. In that case, you need to do the full process described in the book, where you basically get in proper contact with the part in question and address its concerns. (Actually it's a bit more complicated than this, since the IFS model holds that there are many different kinds of parts that may have relationships with each other - so the "beer-drinking part" may be drinking beer in order to keep the traumatized part numb and safely out of consciousness, so you may actually need to deal with two different parts separately. The book goes into a lot more detail.)
Shared publiclyView activity