When you join the CBT club, the first thing they teach you is that you're wrong about everything. It isn't that everybody hates you and you'll be alone forever because of that embarrassing snafu at the holiday party. It's that you're over-generalizing from a single incident, you're mind-reading other people's thoughts without evidence, and you just haven't learned how to be kind to yourself yet.

They teach you that how you feel is a result of how you think, and not the other way around. They teach you to trust your thoughts less. They teach you to notice the self-critical voice inside your head and question it; find all the ways in which it's wrong. A favorite CBT exercise is the two-column technique, where on the left side of the page you write down what your inner critic says, and then on the right side you write a more self-compassionate rebuttal, like what you would say to a friend if they felt that way. Left side: I suck at dating. Right side: I've had some good dates, and I deserve credit for trying. Left side: The boss hated my presentation. Right side: I don't know that, since I haven't heard any feedback one way or another. I'll just ask about it tomorrow.

It's like a perpetual game of Gollum vs Smeagol, where the goal is to understand that Gollum does not have a monopoly on the truth, and in fact his views are often unjustifiably bleak. Then you gradually develop Smeagol's voice as a kinder and more realistic replacement, until it becomes effortless and automatic, and Gollum gives up in disgust and away he goes, precious.

This works well for statements that are easily falsifiable, and where Smeagol can confidently say that his statement is closer to the truth. Gollum: I am a failure. Smeagol: Clearly I don't fail at everything; I successfully brushed my teeth this morning. (Someone please draw me a picture of Smeagol brushing his teeth; I will love you forever.) Gollum: I'll never find another creature to make little Gollums with. Smeagol: I may find such a creature -- nothing in the future has probability 0% or 100%, so I can't say that I never will.

This more-true-than view of CBT is hard to argue against, at least if you see truth as a self-evident good thing. But it hits a limit with statements whose truth is more difficult to determine. Consider the pair of statements: I will [probably / probably not] find a suitable life partner. Figuring out which one of these is more true requires picking a frame of reference. If my frame of reference is my own life experience so far and nothing else, then the answer is no -- I haven't found someone yet. If my frame of reference is all people alive today, then the answer is yes -- most people eventually find someone. In between these extremes are a myriad reference points that try to incorporate some prior information about my personality and preferences, then look at statistics about other people who fall into the same bucket (MBTI type, age at first relationship, breadth of social network, etc). Which of these reference points is more true?

When the more-true-than tower falls, CBT retreats one level below, into the more-constructive-than fortress. Even when I can't determine which statement is more true, I can see that believing one or the other will lead to very different behavior on my part. And some of those behaviors might take me closer or farther away from the outcomes that I want. In this view of CBT, I choose what I believe not based on what is more true (undecidable), but based on what is more self-compassionate and more optimistic and more likely to get me what I want. This is clearly a less comfortable rule of thumb for choosing my beliefs than the truth-based approach -- see road to hell, good intentions, etc. But it still makes sense if I optimize for my own happiness, and the only person responsible for my happiness is me.

Constructing counter-arguments in your head is a powerful piece of mental software, and like all tools, it can be used for good and for evil. Underneath the more-constructive-than fortress lies yet another layer, the rationalize-anything basement. Here the truth heuristic doesn't help, because you're considering possibilities / future paths / different branches of the fig tree, neither more true or false than the others. The constructiveness heuristic also fails, because you're not even sure of what you want. Or worse, in the true CBT spirit of questioning everything, you doubt the validity of what you think you want. Maybe you want to be a rock star but hate practicing the guitar. Maybe you want to be a writer but you don't enjoy the process of writing. Maybe you're in limerence with the idea of someone but not the person herself.

And so you start dismantling your own dreams, scrutinizing every piece, looking for counter-arguments using exactly the same techniques that they taught you for fighting Gollum. Left side: I dream of my own house in Portland with deep greenery in every window and raspberries growing in the backyard. Right side: Yeah but renting is more financially efficient. My friends are here and I don't know anyone in Portland. (Cheryl Strayed doesn't count.) The jobs here pay much better. There's good biking weather for ten months out of every year. And I can buy raspberries at the farmers' market. Maybe I should just decorate my room a little and stop daydreaming about greener grass.

In this way you can convince yourself to be content with your current situation. You can talk yourself out of any "dreams," because why give up something good and familiar in favor of a big unknown? Something about this seems wrong to me. How do you distinguish between dreams that are worth throwing everything away for, and dreams that are better left as fantasies? I'm still looking for a good heuristic.